Method for evaluating and predicting fishy smell of kelp

Abstract

The invention provides a kelp fishy smell evaluation and prediction method, which comprises the following four steps: step 1, carrying out HS-SPME-based pretreatment on a kelp sample, which sequentially comprises four processes of extraction, enrichment, concentration and desorption; step 2, collecting GC-MS spectrogram information of different kelp samples; step 3, performing retrieval qualitative and quantitative analysis on components in GC-MS maps of kelp samples in different production areas according to a standard mass spectrogram library (NISTWEILY), extracting common information of sample components from different sources, calculating OAV values of common chemical components, and performing related odor description; and 4, screening key fishy smell substances in the kelp, evaluating and determining the contribution degree of the fishy smell substances of the kelp by adopting principal component analysis and entropy weight analysis, and constructing a prediction model based on a random forest machine learning algorithm. Therefore, the method has the advantages that volatile flavor substances of the kelp in different regions are deeply analyzed, qualitative and quantitative analysis on the substances is realized, and the characteristic fishy smell components in the kelp are successfully identified.

Description

Technical Field

[0001] The invention belongs to the field of deep processing of aquatic products, relates to quality control, flavor analysis and processing technology of aquatic products, and particularly relates to an evaluation and prediction method for the fishy smell of kelp. Background Art

[0002] my country ranks first in the world in kelp production and is a major producer and user of kelp resources. Kelp, also known as kelp, kelp and Jiangbaicai, grows in low-temperature seawater and is rich in polysaccharides, inorganic salts, proteins, amino acids, carotene, etc. It has the characteristics of low energy and low fat. It also has multiple functions such as anti-oxidation, lowering blood pressure and lowering blood lipids. It plays an important role in human health and longevity. At the same time, it enjoys the title of "longevity vegetable" in Japan. Based on researchers' research on special nutrients in kelp, such as polysaccharides and polyphenols, the value of kelp has been recognized and valued by people, and has gradually become an important product of marine algae development. Therefore, kelp is one of the most important economic algae in my country.

[0003] Kelp has different degrees of fishy smell due to its unique growth environment. This fishy smell is closely related to specific compounds in the volatile components of kelp. The composition of volatile components in kelp is complex and diverse, but the specific compounds that contribute to the fishy smell have not yet been clarified. The fishy smell has a significant negative impact on the taste, flavor and consumer acceptance of kelp, reducing the edible quality of kelp, resulting in a decrease in the acceptance of kelp by some consumers, which in turn affects the market demand. Current research on the fishy smell of kelp mainly focuses on evaluating the degree of fishy smell of kelp through sensory scoring. However, sensory evaluation is easily affected by subjective factors of the evaluators and lacks objectivity and consistency. Therefore, developing an objective and accurate method for evaluating the fishy smell of kelp is of great significance for improving the edible quality of kelp and expanding market consumption.

[0004] There is no detailed report on the identification of kelp fishy substances. Therefore, it is necessary to research and develop a technical method that is easy to operate and can accurately and quickly evaluate and predict the degree of kelp fishy smell based on the existing technical level.

[0005] Gas chromatography-mass spectrometry (GC-MS) is a major breakthrough in the field of instrumental analytical science. It combines efficient separation technology with mass spectrometry. This technology is favored for its advantages such as fast detection speed and low solvent consumption, especially in qualitative and quantitative analysis, showing the characteristics of fast, accurate and high sensitivity. GC-MS is mainly used for the analysis of volatile and semi-volatile organic compounds in complex samples, and it has broad application prospects in the analysis of volatile substances. Through precise separation and sensitive detection, GC-MS technology can reveal detailed information on trace or even trace components in samples, providing a powerful analytical tool for many fields such as environmental monitoring, food safety, and flavor and fragrance analysis. Summary of the invention

[0006] The invention proposes a method for evaluating and predicting the fishy smell of kelp, which solves the problems in the prior art of analysis and identification of key fishy smell components of kelp, scientific evaluation and prediction of fishy smell removal effect, optimization of fishy smell removal process and analysis of changes in flavor substances after fishy smell removal.

[0007] The technical solution of the present invention provides a method for evaluating and predicting the fishy smell of kelp, comprising four steps:

[0008] Step 1: Pre-treatment of kelp samples based on HS-SPME, including four processes of extraction, enrichment, concentration and desorption;

[0009] Step 2, collecting GC-MS spectrum information of different kelp samples;

[0010] Step 3: According to the standard mass spectrum library (NIST\WEILY), the components in the GC-MS spectra of kelp samples from different production areas were searched for qualitative and quantitative analysis, the common information of the components of samples from different sources was extracted, the OAV values ​​of the common chemical components were calculated, and the relevant odor description was performed;

[0011]

[0012] Where: Ax, A TMP are the peak areas of the volatile substance to be measured and the internal standard TMP, respectively; C TMP is the mass concentration of added TMP; V TMP is the volume of TMP added; m k For the mass of kelp sample added.

[0013]

[0014] Where: C i and OT i They are the content of volatile i / (ng / g) and the odor threshold / (ug / kg) respectively. Compounds with OAV greater than 1 have odor activity and can be defined as odor active substances.

[0015] Step 4: Screen the key fishy substances in kelp and use principal component analysis and entropy weight analysis to evaluate and determine the contribution of kelp fishy substances.

[0016] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, in step one, the specific steps include: weighing a certain mass of kelp into a headspace bottle, adding an internal standard solution (TMP) of a certain concentration, preheating for a period of time t at a certain temperature T, pushing the SPME handle to extend the extraction fiber into the headspace bottle, and then placing it in a magnetic stirring heater, staying for a certain period of time t1 at a fixed equilibrium temperature T1, retracting the extraction fiber head, taking out the SPME handle from the headspace bottle, inserting it into the GC-MS injection port and analyzing for a period of time t2.

[0017] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, in step 2, the chromatographic conditions for collecting the GC-MS spectrum information of the kelp sample are as follows: (1) a 30m×0.25mm×0.25μm DB-WAX quartz capillary column; (2) a heating program: an initial temperature of 35°C, maintained for 2min, raised to 180°C at a rate of 5°C / min, maintained for 3min, and finally raised to 230°C at a rate of 10°C / min, maintained for 2min.; (3) a flow rate of He carrier gas of 0.8-1.0mL / min. The mass spectrometry conditions for collecting the GC-MS spectrum information of the kelp sample are as follows: an EI ion source, an electron energy of 70eV, an ion source temperature of 230°C, a quadrupole temperature of 150°C, and a collection mass range of 35-550m / z.

[0018] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, step four is divided into the following operations: (a) constructing a fishy smell evaluation model based on principal component analysis; (b) constructing a fishy smell evaluation model based on entropy weight analysis; (c) sensory experiment verification and association analysis; (d) constructing a fishy smell prediction model.

[0019] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, the extraction fiber head adopts a 50 / 30 μm DVB / CAR / PDMS fiber head, and the extraction fiber head needs to be aged at 250° C. before use until the baseline of the chromatogram is stable and free of impurities.

[0020] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, the preheating temperature T is 60-70°C, the equilibrium temperature T1 is 60-70°C, the time t is 5-10min, t1 is 30-40min, and t2 is 5min.

[0021] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, the operation process of constructing the fishy smell evaluation model based on principal component analysis is: 1. standardizing the original data; 2. performing matrix analysis of each component based on the OVA value of the key fishy smell component, and analyzing the eigenvalues ​​and contribution rates of the correlation matrix; 3. determining the index weights of various key fishy smell components based on the principal component scores; multiplying the content of the key fishy smell component by the corresponding index weight value, and then summing them to obtain the fishy smell value assigned to the sample by the model, and taking the average value after multiple experiments as the final sample fishy smell assignment, referred to as the "fishy smell value".

[0022] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, the operation process of constructing the fishy smell evaluation model based on entropy weight analysis is: ① establish the original data matrix and normalize the index; ② determine the information entropy value of each index by the OVA value of the key fishy smell component; ③ determine the weight of various key fishy smell component indexes by the difference coefficient. The content of the key fishy smell component is multiplied by the corresponding index weight value, and then the sum is obtained to obtain the fishy smell assignment of the sample by the model, and the average value is taken as the final sample fishy smell assignment after multiple tests, which is referred to as the "fishy smell value".

[0023] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, the operation process of the sensory experiment verification and correlation analysis is: using a quantitative description sensory evaluation method, inviting 10 members who have long studied kelp to form an evaluation team, weighing 5g of kelp sample and placing it in a 20mL headspace bottle, balancing it in a 70℃ water bath for 15min, referring to GB / T 10220-2012 "General Theory of Sensory Analysis Methodology", sensory evaluation is performed at room temperature, and the evaluation members record the intensity of five odors according to the samples, and perform sensory evaluation on the smell of kelp according to the standards in Table 1. The higher the score, the heavier the taste, and the full score is 5 points. The evaluator needs to rest between evaluating different samples to relieve olfactory fatigue. After the collected sensory descriptors are sorted and summarized, the team members are trained, and a series of different samples are provided to the evaluators. After in-depth discussion, five odor sensory descriptors are finally determined, which are fishy smell, grassy smell, fruity smell, earthy smell and fat smell.

[0024] Table 1. Scoring criteria for sensory evaluation of kelp

[0025]

[0026] The correlation analysis method was used to analyze the correlation between the sample fishy smell values ​​obtained by the two fishy smell evaluation models and the corresponding fishy smell scores, thus obtaining the correlation coefficients between the two fishy smell evaluation models and sensory evaluation.

[0027] According to a method for evaluating and predicting the fishy smell of kelp of the present invention, the operation process of constructing the fishy smell prediction model is: by collecting different kelp samples for HS-SPME-GC-MS detection and analysis and fishy smell sensory scoring, the OVA values ​​of key fishy smell substances detected by different kelp samples and their corresponding sensory fishy smell scores are combined to build a kelp key fishy smell substance database. The kelp fishy smell evaluation model is constructed by principal component analysis and entropy weight analysis respectively, and the prediction model is constructed by random forest machine learning algorithm, and the fishy smell score of the kelp is predicted by using the key fishy smell substances in the kelp.

[0028] The kelp odor evaluation model has a high correlation with the sensory experiment verification results, R 2 >0.65(p<0.001), the established model was accurate and reliable.

[0029] After adopting the above technical scheme, the beneficial effects of the present invention are as follows: This study uses HS-SPME-GC-MS technology to conduct an in-depth analysis of the volatile flavor substances of kelp in different regions, and realizes the qualitative and quantitative analysis of these substances. Combining the odor activity value (OAV) method and the odor description method, the characteristic fishy components in kelp were successfully identified. In addition, this study also constructed a scientific and objective kelp fishy smell evaluation model, determined the contribution of key characteristic fishy substances to the overall fishy smell, and verified the reliability of the model through sensory evaluation. Furthermore, a kelp fishy smell prediction model was constructed using a random forest machine learning algorithm, which not only expands people's understanding of the formation mechanism of kelp fishy smell, but also provides a theoretical basis and technical support for the study of kelp fishy smell. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0031] Figure 1 It is an overview diagram of the technical flow chart of the present invention;

[0032] Figure 2 This is a graph showing the GC-MS analysis results of kelp samples from five different production areas of the present invention;

[0033] Figure 3 is a graph of volatile substance content of kelp samples from five different production areas of the present invention;

[0034] Figure 4 This is a heat map of OAV values ​​of volatile flavor substances of kelp from different regions of the present invention;

[0035] Figure 5 A diagram of a kelp odor evaluation model constructed by principal component analysis of the present invention;

[0036] Figure 6 A diagram of a kelp odor evaluation model constructed by entropy value weight analysis of the present invention;

[0037] Figure 7 It is a correlation analysis diagram of the two fishy smell evaluation model scores and the fishy smell sensory scores of the present invention;

[0038] Figure 8 A schematic diagram of a kelp odor prediction model constructed based on a random forest machine learning algorithm of the present invention;

[0039] Fig. 9 This is a sensory score result diagram of the fishy smell prediction of kelp samples based on the constructed random forest fishy smell prediction model of the present invention;

[0040] Fig.10 This is a graph showing the effect of tea polyphenols concentration on the fishy smell of kelp according to the present invention;

[0041] Fig.11 It is a correlation analysis diagram between the main component fishy smell evaluation model score of the present invention and the fishy smell sensory score;

[0042] Fig.12 This is a graph showing the effect of yeast fermentation concentration on the degree of kelp fishy smell of the present invention;

[0043] Fig.13 It is a correlation analysis diagram between the main component fishy smell evaluation model score of the present invention and the fishy smell sensory score;

[0044] Fig.14 This is a graph showing the effect of the yeast fermentation temperature of the present invention on the degree of kelp fishy smell;

[0045] Fig.15 It is a correlation analysis diagram between the main component fishy smell evaluation model score of the present invention and the fishy smell sensory score. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] The present invention provides a method for evaluating and predicting the fishy smell of kelp. In this embodiment, kelp samples from different coastal production areas across the country are selected as research objects.

[0048] Step 1: Perform HS-SPME pretreatment on kelp samples from different production areas.

[0049] Take 3 groups of kelp samples from different production areas, crush them, take about 5 g of each and put them in a 20 mL headspace bottle, add a certain volume of internal standard solution, preheat at 60-70 °C for 5-10 min, push the SPME handle to extend the extraction fiber head into the headspace bottle, and then place it on a magnetic stirring heater with an equilibrium temperature of 60-70 °C. After keeping it for 30-40 min, retract the extraction fiber head, take out the solid phase handle, insert it into the injection port of GC-MS, and desorb at the gas chromatography injection port for 5 min.

[0050] Step 2: Collect GC-MS spectrum information of kelp samples from different production areas.

[0051] Set the parameters of the gas phase and mass spectrometer. The specific parameters are as follows:

[0052] Chromatographic conditions: (1) 30 m × 0.25 mm × 0.25 μm DB-WAX quartz capillary column; (2) Heating program: initial temperature 35 °C, hold for 2 min, increase to 180 °C at a rate of 5 °C / min, hold for 3 min, and finally increase to 230 °C at a rate of 10 °C / min, hold for 2 min; (3) The flow rate of He carrier gas was 0.8-1.0 mL / min.

[0053] Mass spectrometry conditions: EI ion source, electron energy 70 eV, ion source temperature 230 °C, quadrupole temperature 150 °C, collection mass range 35-550 m / z.

[0054] After the instrument is stable, the series of samples obtained through pre-treatment in step 1 are manually injected and analyzed.

[0055] Step 3: Use the GC-MS analysis program to obtain the volatile component spectrum data, analyze the GC-MS spectrum information of all samples, and obtain the results, including the following steps:

[0056] The components of the GC-MS spectrum were searched and qualitatively identified based on the NIST and standard mass spectrum libraries, and quantitatively analyzed to calculate the OAV values ​​of the common chemical components and query the related odor descriptions.

[0057]

[0058] Where: Ax, A TMP are the peak areas of the volatile substance to be measured and the internal standard TMP, respectively; C TMP is the mass concentration of added TMP; V TMP is the volume of TMP added; m k For the mass of kelp sample added.

[0059]

[0060] Where: C iand OT i They are the content of volatile i / (ng / g) and the odor threshold / (ug / kg) respectively. Compounds with OAV greater than 1 have odor activity and can be defined as odor active substances.

[0061] Step 4: Select key fishy odor substances with larger OAV values ​​in kelp, and use principal component analysis (1) or entropy weight analysis (2) to determine the contribution (weight ratio) of the kelp fishy odor substances and establish the corresponding model.

[0062] (a) Construction of fishy odor evaluation model based on principal component analysis: ① Standardize the original data; ② Perform component matrix analysis based on the OVA value of key fishy odor components, and analyze the eigenvalues ​​and contribution rates of the correlation matrix; ③ Determine the index weights of various key fishy odor components based on the principal component scores; The model assigns fishy odor values ​​to samples by multiplying the content of key fishy odor components by the corresponding index weight values, and then summing them up. After multiple tests, the average value is taken as the final sample fishy odor assignment, which is referred to as "fishy odor value".

[0063] (b) Construction of fishy smell evaluation model based on entropy weight analysis: ① Establish the original data matrix and normalize the indicators; ② Determine the information entropy value of each indicator by the OVA value of the key fishy smell component; ③ Determine the weight of various key fishy smell component indicators by the difference coefficient. The model assigns the fishy smell of the sample by multiplying the content of the key fishy smell component and the corresponding indicator weight value, and then sums them up. After multiple tests, the average value is taken as the final sample fishy smell assignment, which is referred to as the "fishy smell value".

[0064] (c) Sensory experiment verification: Invite experts to form an evaluation team to conduct sensory evaluation and scoring of different kelp samples.

[0065] The quantitative descriptive sensory evaluation method was adopted. Ten members who had been studying kelp for a long time were invited to form an evaluation team. 5g of kelp sample was weighed and placed in a 20mL headspace bottle. The sample was balanced in a 70℃ water bath for 15min. The sensory evaluation was carried out at room temperature according to GB / T10220-2012 "General Theory of Sensory Analysis Methodology". The evaluation members recorded the intensity of five odors according to the samples. The sensory evaluation of the odor of kelp was carried out according to the standards in Table 1. The higher the score, the stronger the smell. The full score is 5 points. The evaluators need to rest between the evaluation of different samples to relieve olfactory fatigue. After the collected sensory descriptors were sorted and summarized, the team members were trained and provided with a series of different samples. After in-depth discussion, five odor sensory descriptors were finally determined, namely fishy, ​​grassy, ​​fruity, earthy and fat.

[0066] Table 1. Scoring criteria for sensory evaluation of kelp

[0067]

[0068] The above model is used to assign fishy smell values ​​to samples and conduct correlation analysis with human sensory evaluation experimental data.

[0069] The correlation analysis method was used to analyze the fishy smell values ​​obtained by the two fishy smell evaluation models and the fishy smell sensory score results of the corresponding samples. The Pearson correlation coefficient was calculated and statistical tests were performed to obtain the correlation between the two fishy smell evaluation models and the sensory evaluation. When p < 0.05, the correlation was judged to be significant.

[0070] (d) Prediction model construction based on random forest algorithm

[0071] By collecting different kelp samples for HS-SPME-GC-MS detection and analysis and fishy sensory scoring, a database of key fishy substances in kelp was established. The OVA value of the key fishy substances in kelp was used as input, and the fishy sensory score was used as output, and a prediction model was constructed using the random forest algorithm. The random forest algorithm extracts samples from the original data set using bootstrap sampling to form multiple sub-datasets and construct a decision tree for each, and finally averages them to obtain the final prediction. The parameters are adjusted according to the number of randomly selected features (mtry) and the number of decision trees (ntree) that constitute the forest, and the model is built using the optimal parameters.

[0072] The random forest machine learning model was used as the prediction model, and the key fishy substances in kelp were used to predict and score its fishy sensory properties.

[0073] Based on the prediction model constructed in step six, the OVA value of the key fishy odor substances of the newly collected kelp samples was input to predict its fishy odor sensory score, and it was compared with the fishy odor sensory score to verify the reliability of the model.

[0074] Example 1

[0075] A method for evaluating and predicting the fishy smell of kelp comprises the following steps (three groups of the same batch are processed and tested):

[0076] (1) Pretreatment of kelp samples based on HS-SPME: Five different products of light dried kelp were selected as raw materials. Then the test was carried out: 5g of kelp sample was weighed into a 20mL headspace bottle, internal standard solution was added, preheated at 70℃ for 5min, the SPME handle was pushed to extend the extraction fiber head into the headspace bottle, and then placed on a magnetic stirring heater with an equilibrium temperature of 70℃, kept for 30min, and then the extraction fiber head was retracted, the solid phase handle was taken out, and inserted into the injection port of GC-MS, and desorbed at the gas chromatography injection port for 5min.

[0077] (2) Collect GC-MS spectrum information of other groups separately. The spectrum results are shown in Figure 2 .

[0078] (3) The GC-MS analysis program was used to obtain the volatile component spectrum data. The GC-MS spectrum information of all samples was analyzed to obtain the results. The volatile component content was as follows: Figure 3 shown.

[0079] (4) Calculate the OAV value of kelp and query the relevant odor description to screen out the key odor substances in kelp, as shown in Table 2 and Figure 4 As shown in the figure, principal component analysis and entropy weight analysis were used to determine the weights of the fishy substances in kelp, and the ones with greater correlation were selected as the kelp fishy smell evaluation model in combination with sensory evaluation, as shown in the figure. Figure 5-Figure 7 Shown

[0080] (5) The random forest machine learning method was used to predict the fishy smell score based on the key fishy smell substances of kelp, and the fishy smell evaluation model of kelp was verified. Figure 8 and Fig. 9 shown.

[0081] Table 2 Thresholds and odor descriptions of key volatile substances in kelp

[0082]

[0083] Example 2

[0084] A method for evaluating and predicting the fishy smell of kelp comprises the following steps (three groups of the same batch are processed and tested):

[0085] (1) Pretreatment of kelp samples based on HS-SPME: Select light dried kelp as raw material, select 15 samples, add different concentrations of tea polyphenol deodorant, the concentrations are 1%, 1.5%, 2%, 2.5%, 3%, and deodorize for a certain time at a certain temperature. Then perform the test: weigh 5g of kelp sample into a 20mL headspace bottle, add internal standard solution, preheat at 70℃ for 5min, push the SPME handle to extend the extraction head fiber head into the headspace bottle, and then place it on a magnetic stirring heater with an equilibrium temperature of 70℃, keep it for 30min, then retract the extraction fiber head, take out the solid phase handle, insert it into the injection port of GC-MS, and desorb it at the gas chromatography injection port for 5min.

[0086] (2) Collect GC-MS spectrum information of other groups separately.

[0087] (3) The volatile component spectrum data were obtained using the GC-MS analysis program, and the content of fishy substances in kelp under the optimal deodorization conditions was obtained, as shown in Table 3.

[0088] (4) The principal component analysis of the kelp odor evaluation model was used to obtain the changes in the kelp odor degree, such as Fig.10 shown.

[0089] (5) Invite sensory evaluation members to conduct sensory evaluation on the deodorized kelp.

[0090] (6) Correlation analysis was performed between the model odor score and the sensory score, such as Fig.11 The correlation coefficient between the model's fishy smell score and the sensory score is 0.8608, indicating that the model is reliable.

[0091] Table 3 Effect of tea polyphenols concentration on the content of kelp odor substances

[0092]

[0093] Example 3

[0094] A method for evaluating and predicting the fishy smell of kelp comprises the following steps (three groups of the same batch are processed and tested):

[0095] (1) Pretreatment of kelp samples based on HS-SPME: Select light dried kelp as raw material, select 15 samples, add different concentrations of yeast fermentation broth, the concentrations are 0.05%, 0.10%, 0.15%, 0.20% and 0.25%, and remove the fishy smell at a certain temperature for a certain time. Then perform the test: weigh 5g of kelp sample into a 20mL headspace bottle, add internal standard solution, preheat at 70℃ for 5min, push the SPME handle to extend the extraction head fiber head into the headspace bottle, and then place it on a magnetic stirring heater with an equilibrium temperature of 70℃, keep it for 30min, then retract the extraction fiber head, take out the solid phase handle, insert it into the injection port of GC-MS, and desorb it at the gas chromatography injection port for 5min.

[0096] (2) Collect GC-MS spectrum information of other groups separately.

[0097] (3) The volatile component spectrum data were obtained using the GC-MS analysis program, and the content of fishy substances in kelp under the optimal deodorization conditions was obtained, as shown in Table 4.

[0098] (4) The principal component analysis of the kelp odor evaluation model was used to obtain the changes in the kelp odor degree, such as Fig.12 shown.

[0099] (5) Invite sensory evaluation members to conduct sensory evaluation on the deodorized kelp.

[0100] (6) Correlation analysis was performed between the model odor score and the sensory score, such as Fig.13 The correlation coefficient between the model's fishy smell score and the sensory score is 0.8608, indicating that the model is reliable.

[0101] Table 4 Effect of yeast concentration on the content of kelp odor substances

[0102]

[0103] Example 4

[0104] A method for evaluating and predicting the fishy smell of kelp comprises the following steps (three groups of the same batch are processed and tested):

[0105] (1) Pretreatment of kelp samples based on HS-SPME: Select light dried kelp as raw material, select 15 samples, add yeast fermentation broth, and the fermentation temperatures are: 24℃, 28℃, 32℃, 36℃ and 40℃. Then test: weigh 5g kelp sample into a 20mL headspace bottle, add internal standard solution, preheat at 70℃ for 5min, push the SPME handle to extend the extraction fiber head into the headspace bottle, and then place it on a magnetic stirring heater with an equilibrium temperature of 70℃, keep it for 30min, then retract the extraction fiber head, take out the solid phase handle, insert it into the injection port of GC-MS, and desorb it at the gas chromatography injection port for 5min.

[0106] (2) Collect GC-MS spectrum information of other groups separately.

[0107] (3) The volatile component spectrum data were obtained using the GC-MS analysis program, and the content of fishy substances in kelp under the optimal deodorization conditions was obtained, as shown in Table 5.

[0108] (4) The principal component analysis of the kelp odor evaluation model was used to obtain the changes in the kelp odor degree, such as Fig.14 shown.

[0109] (5) Invite sensory evaluation members to conduct sensory evaluation on the deodorized kelp.

[0110] (6) Correlation analysis was performed between the model odor score and the sensory score, such as Fig.15 The correlation coefficient between the model's fishy smell score and the sensory score is 0.8608, indicating that the model is reliable.

[0111] The above are only preferred embodiments of the invention, but the design and concept of the present invention are not limited thereto.

[0112] Table 5 Effect of yeast fermentation on the content of kelp odor substances

[0113]

[0114] This study used HS-SPME-GC-MS technology to conduct an in-depth analysis of the volatile flavor substances in kelp from different regions, and achieved qualitative and quantitative analysis of these substances. Combining the odor activity value (OAV) method and the odor description method, the characteristic fishy components in kelp were successfully identified. In addition, this study also constructed a kelp fishy smell evaluation model, determined the contribution of each fishy substance to the overall fishy smell, and verified the reliability of the model through sensory evaluation. Furthermore, a kelp fishy smell prediction model was constructed using the random forest machine learning algorithm, which not only expanded people's understanding of the formation mechanism of kelp fishy smell, but also provided a theoretical basis and technical support for the study of kelp fishy smell.

[0115] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating and predicting the fishy smell of kelp, characterized in that: There are four steps involved: Step 1: Pre-treatment of kelp samples based on HS-SPME, including four processes of extraction, enrichment, concentration and desorption; Step 2, collecting GC-MS spectrum information of different kelp samples; Step 3: Retrieve the components in the GC-MS spectra of kelp samples from different production areas according to the standard mass spectrum library, perform qualitative and quantitative analysis, extract the common information of the components of samples from different sources, calculate the OAV values ​​of the common chemical components and perform relevant odor descriptions; Step 4: Screen the key fishy substances in kelp and use principal component analysis and entropy weight analysis to evaluate and determine the contribution of kelp fishy substances.

2. The method for evaluating and predicting the fishy smell of kelp according to claim 1, characterized in that: In the step one, the specific steps include: weighing a certain mass of kelp into a headspace bottle, adding an internal standard solution of a certain concentration, preheating for a period of time t at a certain temperature T, pushing the SPME handle to extend the extraction fiber into the headspace bottle, and then placing it in a magnetic stirring heater, staying for a certain period of time t1 at a fixed equilibrium temperature T1, retracting the extraction fiber head, taking out the SPME handle from the headspace bottle, inserting it into the GC-MS injection port and analyzing for a period of time t2.

3. The method for evaluating and predicting the fishy smell of kelp according to claim 1, characterized in that: In the step 2, the chromatographic conditions for collecting the GC-MS spectrum information of the kelp sample are as follows: (1) a 30 m×0.25 mm×0.25 µm DB-WAX quartz capillary column; (2) a heating program: an initial temperature of 35°C, maintained for 2 min, increased to 180°C at a rate of 5°C / min, maintained for 3 min, and finally increased to 230°C at a rate of 10°C / min, maintained for 2 min; (3) a He carrier gas flow rate of 0.8-1.0 mL / min, and mass spectrometry conditions for collecting the GC-MS spectrum information of the kelp sample are as follows: an EI ion source, an electron energy of 70 eV, an ion source temperature of 230°C, a quadrupole temperature of 150°C, and a collection mass range of 35-550 m / z.

4. The method for evaluating and predicting the fishy smell of kelp according to claim 1, characterized in that: The step 4 is divided into the following operations: (a) Construction of fishy odor evaluation model based on principal component analysis; (b) Construction of fishy odor evaluation model based on entropy weight analysis; (c) Sensory experiment verification and correlation analysis; (d) Construction of fishy odor prediction model.

5. The method for evaluating and predicting the fishy smell of kelp according to claim 2, characterized in that: The extraction fiber head adopts a 50 / 30 μm DVB / CAR / PDMS fiber head, and the extraction fiber head needs to be aged at a temperature of 250° C. before use until the baseline of the chromatogram tends to be stable and free of impurities.

6. The method for evaluating and predicting the fishy smell of kelp according to claim 2, characterized in that: The preheating temperature T is 60-70°C, the equilibrium temperature T1 is 60-70°C, the time t is 5-10 min, t1 is 30-40 min, and t2 is 5 min.

7. The method for evaluating and predicting the fishy smell of kelp according to claim 4, characterized in that: The operation process of constructing the fishy smell evaluation model based on principal component analysis is: ① standardizing the original data; ② performing matrix analysis of each component based on the OVA value of the key fishy smell component, and analyzing the eigenvalue and contribution rate of the correlation matrix; ③ determining the index weights of various key fishy smell components based on the principal component scores; multiplying the content of the key fishy smell component by the corresponding index weight value, and then summing them to obtain the fishy smell value assigned to the sample by the model, and taking the average value after multiple tests as the final sample fishy smell value, referred to as "fishy smell value".

8. The method for evaluating and predicting the fishy smell of kelp according to claim 7, characterized in that: The operation process of constructing the fishy smell evaluation model based on entropy weight analysis is: ① establish the original data matrix and normalize the indicators; ② determine the information entropy value of each indicator based on the OVA value of the key fishy smell component; ③ determine the weight of various key fishy smell component indicators based on the difference coefficient, and assign the fishy smell value of the sample to the model by multiplying the content of the key fishy smell component by the corresponding indicator weight value, and then summing them up. After multiple tests, the average value is taken as the final sample fishy smell assignment, which is referred to as "fishy smell value".

9. The method for evaluating and predicting the fishy smell of kelp according to claim 8, characterized in that: The operational process of the sensory experiment verification and correlation analysis is as follows: a quantitative descriptive sensory evaluation method is used, 10 members who have been studying kelp for a long time are invited to form an evaluation team, 5 g of kelp sample is weighed and placed in a 20 mL headspace bottle, balanced in a 70°C water bath for 15 min, and sensory evaluation is performed at room temperature. The evaluation members record the intensity of five odors based on the samples, and use the correlation analysis method to perform correlation analysis on the sample fishy smell values ​​obtained by the two fishy smell evaluation models and the corresponding fishy smell scores, thereby obtaining the correlation coefficient between the two fishy smell evaluation models and the sensory evaluation.

10. The method for evaluating and predicting the fishy smell of kelp according to claim 9, characterized in that: The operation process of constructing the fishy smell prediction model is as follows: by collecting different kelp samples for HS-SPME-GC-MS detection and analysis and fishy smell sensory scoring, the OVA values ​​of key fishy substances detected in different kelp samples and their corresponding sensory fishy smell score values ​​are combined to build a kelp key fishy smell substance database, principal component analysis and entropy weight analysis are used to respectively construct kelp fishy smell evaluation models, a prediction model is constructed with a random forest machine learning algorithm, and the fishy smell score of the kelp is predicted using the key fishy substances in the kelp.

Images