A quantitative evaluation method for multi-dimensional aroma of tea based on hierarchical analysis

Abstract

A quantitative evaluation method for the multi-dimensional aroma of tea based on hierarchical analysis comprises: 1) collecting and preparing tea samples to construct the multi-dimensional characteristic aroma attributes of tea; 2) using the multi-dimensional characteristic aroma attributes of tea to construct a hierarchical structure model of the multi-dimensional characteristic aroma attributes of tea; 3) calculating the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea based on the hierarchical analysis method; 4) calculating the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea; 5) based on the quantitative descriptive analysis method, having expert evaluators quantitatively evaluate the intensity of the characteristic aroma attributes to obtain the quantitative value of the intensity of the multi-dimensional characteristic aroma attributes of tea; 6) calculating the weighted total score of the aroma characteristics of the sample based on the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea and the quantitative value of the intensity of the multi-dimensional characteristic aroma attributes of tea. The present invention introduces the concept of the multi-dimensional characteristic aroma of tea, and the evaluation of the aroma of tea is more comprehensive.

Description

Technical Field

[0001] The present invention relates to the field of tea evaluation, and in particular to a multi-dimensional tea aroma quantitative evaluation method based on hierarchical analysis. Background Art

[0002] The sensory flavor of tea is a primary concern for consumers when purchasing. Traditional tea evaluation is an expert-led method that ranks tea quality through comprehensive descriptions and overall scores (GB / T 23776-2018). While this method is widely adopted as a standard evaluation technique, it also suffers from the drawback of insufficiently precise quantification of characteristic attributes.

[0003] At present, food sensory evaluation mostly uses quantitative descriptive analysis, which uses standardized testing procedures and professional evaluation teams to quantify the sensory characteristics of products (such as taste, smell, color, texture, etc.). This method has been widely used in various food, beverage and agricultural products. Quantitative descriptive analysis technology is also used in the field of tea, but it is currently mainly used for the aroma of the tea leaves after brewing, and the evaluation dimension is single. The perception of tea aroma is a multidimensional system, including the aroma of the tea leaves, the aroma of the soup surface and the aroma after the nose, all three of which will affect consumer preferences. Therefore, it is necessary to establish a multi-dimensional sensory evaluation method for tea to form a more comprehensive quality evaluation result.

[0004] Multidimensional quantitative descriptive analysis generates a series of quantitative data for the various sensory attributes of tea. The key to ultimately applying this method to tea evaluation is how to develop a scoring system based on this data. Therefore, it is necessary to further integrate quantitative evaluation with multidimensional perception to construct a model for calculating the overall evaluation score. The Analytic Hierarchy Process (AHP) decomposes complex, multi-category, and multi-factor problems into distinct factors based on the nature of the problem and the intended objectives. Based on the interrelationships and affiliations between these factors, a multi-level analytical structure is formed, thereby reducing the problem to the weights of the lowest-level sub-indicators relative to the highest-level target indicators. This method has been previously applied in fields such as safety science and environmental science, but its application in the food industry has been limited. Furthermore, the AHP only considers the weights of sub-indicators relative to the highest-level indicators, but does not evaluate the weighting effect of these weights. In food evaluation, the lowest-level sub-indicators (sensory attributes) can have varying preferences for consumers, which should be considered.

[0005] In summary, through the hierarchical analysis-based multi-dimensional characteristic aroma evaluation method of tea established by this patent, its multi-dimensional characteristic aroma attributes (leaf bottom aroma, soup surface aroma, and after-nasal aroma) are quantitatively evaluated. Combined with the preference effect of different characteristic aroma attributes, the weighted total score of the aroma characteristics of the sample is calculated to form a comprehensive quantitative evaluation model for the aroma quality of tea. Summary of the Invention

[0006] The purpose of the present invention is to provide a multi-dimensional aroma quantitative evaluation method for tea based on hierarchical analysis, which can evaluate the aroma quality of tea more comprehensively and accurately from multiple dimensions, so as to solve the existing technical defects and unattainable technical requirements.

[0007] To achieve the above object, the present invention provides the following technical solution: a method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis, comprising the following steps:

[0008] 1) Collect and prepare tea samples according to standards, and determine the multi-dimensional characteristic aroma attributes of tea based on the Chinese tea sensory aroma wheel;

[0009] Multidimensional aroma comprises three components: leaf aroma, tea aroma, and retronasal aroma. Leaf aroma refers to the aroma released by the remaining tea leaves after brewing and filtering out the tea. Tea aroma refers to the aroma released by the filtered tea liquid. Retronasal aroma refers to the aroma released in the mouth and perceived through the nose during sipping. By analyzing these three different dimensions of perception, we can comprehensively describe the aroma characteristics released by tea in different dimensions, thereby providing a more comprehensive quality assessment and description.

[0010] 2) constructing a hierarchical structure model of the multi-dimensional characteristic aroma attributes of tea consisting of multiple levels according to the multi-dimensional characteristic aroma attributes of tea obtained in step 1);

[0011] 3) Based on the hierarchical analysis method, the hierarchical structure model of the multi-dimensional characteristic aroma attributes of tea obtained in step 2) is used to calculate the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea

[0012] 3.1) Construct judgment matrices for different levels in the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea, and have expert evaluators score the importance of each sub-indicator of each level in the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea obtained in step 2), and further normalize the judgment matrix;

[0013] 3.2) Based on the hierarchical analysis method, calculate the weight coefficients of the sub-indicators at each level of the hierarchical structure model of the multi-dimensional characteristic aroma attributes of the leaves in step 2), and then calculate the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of the tea leaves through the weight calculation coefficients of the sub-indicators at each level;

[0014] 3.3) Based on the consistency matrix method, the consistency of the importance scores obtained by the expert evaluators was evaluated;

[0015] 4) The preference effect is weighted for the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea obtained in step 3), and further, the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea is calculated.

[0016] 4.1) Using the multi-dimensional characteristic aroma attributes of tea in step 1), analyze consumers' preferences for different attributes of tea to obtain the preference effect of each attribute;

[0017] 4.2) Based on the preference effect, the preference effect is assigned to the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of the tea obtained in step 3), and then the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of the tea is calculated based on the weighted result;

[0018] V) Based on the quantitative descriptive analysis method, a reference sample of multi-dimensional characteristic aroma attributes is established, tea samples are prepared and presented, and more than 10 expert evaluators are organized to quantitatively evaluate the intensity of the characteristic aroma attributes of tea samples in three dimensions: leaf aroma, soup aroma, and after-nose aroma, and obtain the quantitative value S of the intensity of the multi-dimensional characteristic aroma attributes of tea. i ;

[0019] 6) Combining the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea after weighting in step 4) and the quantitative value of the intensity of the multi-dimensional characteristic aroma attribute of tea in step 5), the weighted total score of the aroma characteristics of the sample is calculated to form a comprehensive quantitative evaluation model of the aroma quality of tea.

[0020] Preferably, in step 1), the specific contents of preparing and presenting the tea sample are:

[0021] 1.1) According to the standard (GB / T 23776-2018), weigh 3.0g of the collected tea leaves and put them into a standard evaluation cup. Pour in 150mL of boiling water and brew for 4 minutes.

[0022] 1.2) Separate the brewed sample tea liquid, leave the tea leaves in the standard evaluation cup for testing, and pour the filtered tea into the standard evaluation bowl for testing.

[0023] 1.3) Organize more than 10 expert evaluators, with an even number of expert evaluators, and divide all expert evaluators into multiple evaluation groups, each of which contains two expert evaluators;

[0024] 1.4) Randomly code the standard evaluation cup containing tea leaves and the standard evaluation bowl containing tea soup. Present each pair of randomly coded standard evaluation cups and standard evaluation bowls to two expert evaluators from the same group. The two expert evaluators will evaluate the aroma of the tea leaves in the standard evaluation cups and the aroma and after-nose aroma of the tea soup in the standard evaluation bowls. The other evaluation groups will follow the same procedure.

[0025] 1.5) Prepare the same tea sample according to 1.1) and 1.2) again. According to 1.4), present a pair of randomly coded standard evaluation cups and standard evaluation bowls to two expert evaluators in the same group. The two expert evaluators exchange the evaluation objects and conduct aroma evaluation.

[0026] Preferably, in step 1), the method for determining the multi-dimensional characteristic aroma attributes of tea is: according to the national standard GB / T 39625-2020, which includes a combination of two methods: the empirical method (based on the tea aroma wheel) and the group discussion method (generated by on-site discussion of the evaluation group).

[0027] 1.6) After evaluation by expert evaluators, the consensus and non-consensus attributes of the multi-dimensional characteristic aroma of the sample are confirmed. Non-consensus attributes should be repeatedly evaluated and opinions solicited through multiple rounds of samples until all attributes are consensus attributes and the final evaluation results are obtained;

[0028] Among them, consensus attributes are attributes on which expert evaluators directly reach consensus, and non-consensus attributes are attributes on which no consensus is directly reached:

[0029] 1.7) Generate a list of multi-dimensional characteristic aroma attributes of tea based on the final results of the expert evaluators.

[0030] In this embodiment, it should be noted that in step 1.5), a pair of tea samples of the same category is replaced, that is, a group of leaf tea samples and tea soup samples are replaced, and then this pair of new tea sample groups are randomly coded. In addition, in step 1.5), the meaning of exchanging the evaluation objects can be understood as: the expert evaluator who first evaluated the tea soup evaluates the leaf tea samples in this step, and the expert evaluator who first evaluated the leaf tea samples evaluates the tea soup samples in this step. Because the tea samples have cooled after one evaluation process, a pair of tea samples belonging to the same category is replaced for evaluation in this step.

[0031] In addition, when expressing their evaluation, expert evaluators refer to GB / T 16861-1997 "Sensory Analysis - Identification and Selection of Descriptors for Establishing Sensory Profiles by Multivariate Analysis Methods", the Tea Sensory Aroma Wheel in T / CTSS 58-2022 "Tea Sensory Flavour Wheel" and GB / T 14487-2017 "Tea Sensory Evaluation Terminology".

[0032] Preferably, the specific content of the step 2) is: the layers are divided into a core layer (tea aroma), a perception layer (specific dimensions of perception: including leaf base aroma, soup surface aroma, and nose aroma), a classification layer (the specific flavor attributes of the tea obtained in step 1 are classified according to their flavor types: such as overall intensity, herbal, fruity, nectar aroma, etc.) and a characteristic layer (the specific flavor attributes of the tea obtained in step 1, such as green aroma, fresh fragrance, tender fragrance, etc.).

[0033] Preferably, the content of step 3.1 is specifically:

[0034] 3.1.1) By comparing the importance of each sub-indicator at each level, refer to the following table to determine the score of the comparison result of the importance of sub-indicator i and sub-indicator j as a ij , based on the score a of the comparison results of the importance between indicators ij Generate judgment matrix A ij , where i and j represent two different sub-indicators, i, j = 1, 2, ..., n, and the importance comparison results of sub-indicators at each level are a ij The reference standards are as follows:

[0035]

[0036] 3.1.2) For the judgment matrix A ij Perform normalization processing to obtain a normalized judgment matrix) Where n is the matrix order.

[0037] Preferably, the specific content of step 3.2) is:

[0038] 3.2.1) Calculate the weight coefficients of the sub-indicators at each level: Add row by row to get the eigenvector Then the eigenvector Perform normalization to obtain the index A in the single ranking of each level i Weight coefficient N for the superior indicator i ,

[0039] 3.2.2) Calculation of the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea:

[0040] Comprehensive weight coefficient of tea's multi-dimensional aroma attributes That is, the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea is calculated through the multi-level product of the sub-weight coefficients.

[0041] Preferably, the specific content of step 3.3) is:

[0042] 3.3.1) For the maximum characteristic root λ of the judgment matrix max Perform the calculation:

[0043] 3.3.2) Further, the matrix consistency index CI is calculated:

[0044] 3.3.3) Further, the consistency ratio value CR is calculated: Among them, RI is the random consistency index. If CR is less than 0.1, the matrix is ​​considered to have good consistency. It is used to judge whether the consistency of the scoring results among expert evaluators is good. If it is good, the obtained weight coefficient is reliable.

[0045] Preferably, the specific steps in step 4) are:

[0046] The specific contents of step 4.1) are:

[0047] 4.1.1) Based on the multi-dimensional characteristic aroma attributes of tea in step 1), survey consumers on their preferences for different attributes, evaluate the survey results, and obtain the preference effects of specific attributes at the characteristic level;

[0048] 4.1.2) The specific criteria for distinguishing preference effects are as follows: a positive effect is defined as a percentage of consumers liking an attribute in the multi-dimensional aroma attribute exceeding 70%, a negative effect is defined as a percentage of consumers liking an attribute below 30%, and a neutral effect is defined as a percentage of consumers liking an attribute between 30% and 70%.

[0049] That is, it can be understood that according to the preferences of consumers, it is possible to judge whether the attributes in the multi-dimensional characteristic aroma attributes have positive effects, negative effects, or neutral effects.

[0050] The specific contents of step 4.2) are:

[0051] M i =P i ×w, where M i is the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea, where the weighting coefficient w of the positive effect characteristic layer sub-indicator is 1, the weighting coefficient w of the negative effect characteristic layer sub-indicator is -1, and the weighting coefficient w of the neutral effect characteristic layer sub-indicator is 0.

[0052] Preferably, in the step 5), a multidimensional aroma evaluation method is proposed based on the quantitative description analysis method using the multidimensional characteristic aroma attributes obtained in step 1), which specifically comprises: establishing a reference sample of multidimensional characteristic aroma attributes, preparing and presenting tea samples, and organizing more than 10 expert evaluators to evaluate the specific attributes of leaf bottom aroma, soup surface aroma and after-nasal aroma in turn, to obtain the quantitative value S of the intensity of the multidimensional characteristic aroma attributes of tea. i ,The evaluation results were recorded and counted using standardized sensory software.

[0053] In this application, in order to prevent fatigue of expert evaluators, the duration of a single evaluation is 3 hours, with a 15-minute break between each sample, and clean water and soda crackers are provided for oral cleaning and sensory recovery to eliminate the impact of the previous round of samples.

[0054] Preferably, in step 6), based on the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea obtained in step 4) and the quantitative value of the intensity of the multi-dimensional characteristic aroma attribute of tea obtained in step 5), a comprehensive quantitative evaluation model of the aroma quality of tea is formed, and the final score K=∑(M i ×S i ).

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] 1. This invention introduces the concept of multi-dimensional characteristic aroma of tea for the first time, and quantitatively evaluates the characteristic aroma attributes of the tea leaves, tea soup and after-nasal aroma after brewing, thereby improving the breadth of the perception level of tea aroma evaluation.

[0057] 2. The present invention combines the preference effect with the hierarchical analysis method to determine the relative importance of different characteristic aroma attributes in the overall aroma score, which can more scientifically and accurately weight the aroma differences between different tea samples.

[0058] 3. The present invention quantitatively describes and analyzes the multi-dimensional characteristic aroma attributes of tea, which is a more accurate quantitative evaluation technology compared to the current national standards. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is the overall logical structure diagram of this application

[0060] Figure 2 Example 1 Longjing tea multi-dimensional characteristic aroma attribute hierarchical structure model

[0061] Figure 3 Quantify the Si values ​​of the 18 multi-dimensional characteristic aroma attributes of Longjing tea (West Lake production area) in Example 1

[0062] Figure 4 Quantify the Si values ​​of the 18 multi-dimensional characteristic aroma attributes of Longjing tea (Qiantang production area) in Example 1

[0063] Figure 5 Quantify the Si values ​​of the 18 multi-dimensional characteristic aroma attributes of Longjing tea (Yuezhou production area) in Example 1 DETAILED DESCRIPTION

[0064] The following is a combination of the embodiments of the present invention Figure 1-5 The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0065] See also Figure 1-5 , embodiments of the present invention:

[0066] Example 1:

[0067] like Figure 1-5 As shown: The present invention is specifically described as follows using Zhejiang Longjing tea as an example:

[0068] First, the relevant instruments used in this example include: The evaluation equipment used in the experiment includes a professional sensory evaluation room, evaluation table, evaluation cups, evaluation bowls, evaluation trays, leaf trays, balances, timers, etc. Experimental data were recorded using the FIZZ Analysis System (Biosystèmes, France), that is, the standardized sensory software is the FIZZ Analysis System (Biosystèmes, France).

[0069] The water used for evaluation was Wahaha (Hangzhou Wahaha Group Co., Ltd., Hangzhou, Zhejiang).

[0070] Specifically: A multi-dimensional aroma quantitative evaluation method for tea based on hierarchical analysis includes the following steps:

[0071] 1) Collect and prepare tea samples according to standards, and determine the multi-dimensional characteristic aroma attributes of tea based on the Chinese tea sensory aroma wheel;

[0072] In the step 1), one Longjing tea sample each from three different Longjing tea producing areas (West Lake, Qiantang District, and Yuezhou District) in Zhejiang Province as specified in GB / T 18650-2008 was collected and stored in a frozen state at -18°C in the dark.

[0073] In step 1), the specific contents of preparing various tea samples according to standard GB / T 23776-2018 are as follows:

[0074] 1.1) Weigh 3.0g of the collected tea leaves into a standard evaluation cup, add 150mL of boiling water, and brew for 4 minutes;

[0075] 1.2) Separate the brewed sample tea liquid, leave the tea leaves in the standard evaluation cup for testing, and pour the filtered tea into the standard evaluation bowl for testing.

[0076] 1.3) Organize 10 expert evaluators and divide all expert evaluators into 5 evaluation groups, each evaluation group contains two expert evaluators;

[0077] 1.4) Randomly code the standard evaluation cup containing tea leaves and the standard evaluation bowl containing tea soup. Present each pair of randomly coded standard evaluation cups and standard evaluation bowls to two expert evaluators from the same group. The two expert evaluators will evaluate the aroma of the tea leaves in the standard evaluation cups and the aroma and after-nose aroma of the tea soup in the standard evaluation bowls. The other evaluation groups will follow the same procedure.

[0078] 1.5) Prepare the same tea sample according to 1.1) and 1.2) again. According to 1.4), present a pair of randomly coded standard evaluation cups and standard evaluation bowls to two expert evaluators in the same group. The two expert evaluators exchange the evaluation objects and conduct aroma evaluation.

[0079] The method for determining the multi-dimensional characteristic aroma attributes of tea is:

[0080] 1.6) After evaluation by expert evaluators, the consensus and non-consensus attributes of the multi-dimensional characteristic aroma of the sample are confirmed. Non-consensus attributes should be repeatedly evaluated and opinions solicited through multiple rounds of samples until all attributes are consensus attributes and the final evaluation results are obtained;

[0081] Among them, consensus attributes are attributes on which expert evaluators directly reach consensus, and non-consensus attributes are attributes on which no consensus is directly reached:;

[0082] 1.7) Generate a list of multi-dimensional characteristic aroma attributes of tea based on the final evaluation results of the expert evaluators, as shown in Table 1

[0083] In this embodiment, it should be noted that the expert evaluators all have skill certificates of senior tea appraisers or above and have a high degree of understanding of the aroma properties of tea;

[0084] When evaluating and presenting the descriptors (common and specific descriptors), expert evaluators refer to GB / T 16861-1997, "Sensory Analysis: Identification and Selection of Descriptors for Establishing Sensory Profiles by Multivariate Analysis," the Tea Sensory Aroma Wheel in T / CTSS 58-2022, "Tea Sensory Flavor Wheel," and GB / T 14487-2017, "Tea Sensory Evaluation Terminology." The descriptor selection process is based on the national standard GB / T 39625-2020 and includes a combination of two methods: an empirical method (based on the Tea Aroma Wheel) and a group discussion method (generated through on-site discussions by the evaluation team). Expert evaluators initially generate sample descriptors based on their experience and sample evaluation results, and further discuss and determine the common and specific descriptors for the samples. Common descriptors are aroma attributes familiar to experts, while specific descriptors require the opinions of all experts in conjunction with the samples until consensus is reached.

[0085] Table 1: List of multi-dimensional characteristic aroma attributes of tea

[0086] serial number Multi-dimensional characteristic aroma properties of Longjing tea definition 1 Overall strength The olfactory intensity of all aroma types mixed 2 Green Air Smells of grass or leaves 3 fresh fragrance Similar to the fresh and pure aroma of green plants 4 Tender dried bamboo shoots The smell of tender bamboo shoots after being dried and cooked 5 Overwintering leaf scent The smell of winter buds in the fields 6 tender and fragrant The pleasant and delicate aroma unique to young tea 7 Lixiang Aroma similar to cooked chestnuts 8 Fresh floral fragrance Fresh flower-like scent 9 sweet floral scent Similar to sweet floral scent 10 Honey fragrance Honey-like aroma 11 milky flavor Dairy-like aroma 12 sweet fragrance sweet aroma

[0087] In this example, the multi-dimensional aroma profile includes quantitative descriptive analysis of three aspects: leaf aroma, tea aroma, and retronasal aroma. Leaf aroma refers to the aroma released by the remaining tea leaves after brewing and filtering the tea; surface aroma refers to the aroma released from the surface of the filtered tea; and retronasal aroma refers to the aroma perceived through the nose during sipping. This systematic sensory evaluation method comprehensively characterizes the aroma characteristics released by tea in different dimensions, providing a more comprehensive quality assessment and description. During actual evaluation, these multi-dimensional aroma analyses are conducted under standardized evaluation conditions to ensure objectivity and comparability of the results.

[0088] like Figure 2 : 2) According to the multi-dimensional characteristic aroma attributes of tea obtained in step 1), a hierarchical model of multi-dimensional characteristic aroma attributes of tea consisting of a multi-level structure is constructed; the hierarchical structure is divided into a core layer, a perception layer, a classification layer and a feature layer.

[0089] 3) Based on the hierarchical analysis method, the hierarchical structure model of the multi-dimensional characteristic aroma attributes of tea obtained in step 2) is used to calculate the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea

[0090] 3.1) Construct judgment matrices for different hierarchical structures in the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea, and have expert evaluators score the importance of each level sub-indicator of each hierarchical structure in the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea obtained in step 2), and further normalize the judgment matrix;

[0091] In this embodiment, specifically, there are 19 expert evaluators (7 men and 12 women) participating in the scoring process in this step, all of whom are from Zhang Yingbin's Tea Sensory Evaluation Innovation Studio of the Tea Research Institute of the Chinese Academy of Agricultural Sciences. The aroma attributes of the Longjing tea selected in the early stage are determined according to the tea sensory aroma wheel in the T / CTSS 58-2022 "Tea Sensory Flavor Wheel" group standard to determine its type, construct a hierarchical structure model of the multi-dimensional characteristic aroma attributes of tea, and establish scoring tables for each level. Among them, for the multi-dimensional characteristic aroma attributes, the weights of the classification layer and the feature layer are consistent; the weights of the perception layer are different.

[0092] 3.1.1) By comparing the importance of each sub-indicator at each level, refer to the following table to determine the score of the comparison result of the importance of sub-indicator i and sub-indicator j as a ij , based on the score a of the comparison results of the importance between indicators ij Generate judgment matrix A ij, where i and j represent two different sub-indicators, i, j = 1, 2, ..., n, and the importance comparison results of sub-indicators at each level are a ij The reference standards are shown in Table 2:

[0093] Table 2: Reference standards for comparison of importance among sub-indicators at each level

[0094]

[0095] 3.1.2) For the judgment matrix A ij Perform normalization processing to obtain a normalized judgment matrix) Where n is the matrix order.

[0096] 3.2) Based on the hierarchical analysis method, calculate the weight coefficients of the sub-indicators at each level of the hierarchical structure model of the multi-dimensional characteristic aroma attributes of the leaves in step 2), and then calculate the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of the tea leaves using the weight coefficients of the sub-indicators at each level;

[0097] 3.2.1) Calculate the weight coefficients of the sub-indicators at each level: Add row by row to get the eigenvector Then the eigenvector Perform normalization to obtain the index A in the single ranking of each level i Weight coefficient N for the superior indicator i ,

[0098] 3.2.2) Calculation of the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea:

[0099] Comprehensive weight coefficient of tea's multi-dimensional aroma attributes The results are shown in Table 3;

[0100] Table 3: Levels and comprehensive weight coefficients of multi-dimensional aroma attributes of tea

[0101]

[0102]

[0103]

[0104] In this embodiment, the hierarchical analysis method decomposes complex multi-category and multi-factor problems into different factors based on the nature of the problem and the expected goals, and forms a multi-level analysis structure model according to the mutual correlation and affiliation between the factors, so that the problem is reduced to the weight value of the lowest level sub-indicator relative to the highest level total indicator, which can be applied within the scope of this patent.

[0105] 3.3) Based on the consistency matrix method, the consistency of the importance scores obtained by the expert evaluators was evaluated;

[0106] 3.3.1) For the maximum characteristic root λ of the judgment matrix max Perform the calculation:

[0107] 3.3.2) Further, the matrix consistency index CI is calculated:

[0108] 3.3.3) Further, the consistency ratio value CR is calculated: Among them, RI is the random consistency index (reference table of random consistency index RI values, such as Table 4). If CR is less than 0.1, it is considered that the matrix has good consistency. It is judged whether the consistency of the scoring results between expert evaluators is good. If it is good, the obtained weight coefficient is reliable. The results are shown in Table 5.

[0109] Table 4 Random Consistency Index RI Value Reference Table:

[0110] Matrix order 1 2 3 4 5 6 7 8 9 10 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49

[0111] Table 5: Consistency ratio values ​​CR of indicators at different levels

[0112]

[0113] Note: The CR values ​​of the consistency evaluation of sub-indicators at different levels are all less than 0.1, indicating that the obtained weight coefficients are reliable.

[0114] 4) Weighting the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea obtained in step 3) by preference effect, and further calculating the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea

[0115] 4.1) Using the multi-dimensional characteristic aroma attributes of tea in step 1), the preference effect of each attribute is obtained based on the consumer's preference for different attributes of tea;

[0116] 4.1.1) Based on the multi-dimensional characteristic aroma attributes of tea in step 1), survey 57 consumers on their preferences for different attributes, evaluate the survey results, and obtain the preference effects of specific attributes at the characteristic level;

[0117] 4.1.2) The specific criteria for distinguishing preference effects are: a consumer preference degree above 70% is a positive effect, a consumer preference degree below 30% is a negative effect, and a consumer preference degree between 30% and 70% is a neutral effect.

[0118] 4.2) Based on the preference effect, the preference effect was assigned to the comprehensive weight coefficients of the multi-dimensional characteristic aroma attributes of tea obtained in step 3) (the weighting results are shown in Table 6). Then, the weighted comprehensive weight coefficients of the multi-dimensional characteristic aroma attributes of tea were calculated based on the weighting results. The results are shown in Table 7.

[0119] M i =P i ×w, where M i is the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea, where the weight coefficient w of the positive effect characteristic layer sub-indicator is 1, the weight coefficient w of the negative effect characteristic layer sub-indicator is -1, and the weight coefficient w of the neutral effect characteristic layer sub-indicator is 0 (as shown in Table 6).

[0120] Table 6: Preference effects of Longjing tea’s multi-dimensional aroma attributes and comprehensive weight coefficients for weighting preference effects

[0121]

[0122] Table 7: Comprehensive weight coefficients of Longjing tea’s multi-dimensional aroma attributes after empowerment

[0123]

[0124]

[0125] like Figure 3-5 As shown in Figure 5), based on the quantitative descriptive analysis method, expert evaluators quantitatively evaluated the intensity of the characteristic aroma attributes of tea samples in three dimensions: leaf aroma, soup aroma, and after-nasal aroma, and obtained the quantitative value of the intensity of the multi-dimensional characteristic aroma attributes of tea;

[0126] Using the multi-dimensional characteristic aroma attributes obtained in step 1), a multi-dimensional aroma evaluation method is proposed based on the quantitative descriptive analysis method. The specific content is: establish a reference sample of multi-dimensional characteristic aroma attributes, prepare and present tea samples, organize 10 expert evaluators to evaluate the specific attributes of leaf bottom aroma, soup surface aroma and after-nasal aroma in turn, and obtain the quantitative value S of the intensity of the multi-dimensional characteristic aroma attributes of tea. i ,The evaluation results were recorded and counted using standardized sensory software;

[0127] Specifically, after completing the sample preparation according to the above method, the evaluation cups and evaluation bowls containing tea leaves and tea soup respectively were randomly coded in turn and presented to 10 expert evaluators (4 males and 6 females, all of whom had professional skills of senior tea appraisers or above, and had received more than 600 hours of tea sensory quantitative description and analysis training in advance). The expert evaluators were divided into 5 groups (2 people in each group), with 1 person in the group evaluating the aroma of the tea leaves and the other person evaluating the aroma of the soup surface and the after-nose aroma. After the evaluation, the same sample was prepared again, and the evaluation objects were exchanged within the group. The sample evaluation was based on the quantitative descriptive analysis method, using the 9-point scaling method to quantitatively evaluate the multi-dimensional characteristic aroma attributes of Longjing tea. In order to prevent expert fatigue, a single evaluation lasted 3 hours, with a 15-minute break between each sample, and water and soda crackers were provided for oral cleaning and sensory recovery to eliminate the impact of the previous round of samples.

[0128] When expert evaluators conducted sensory evaluations on Longjing tea test samples, they were provided with a reference sample table of Longjing tea’s multi-dimensional characteristic aroma attributes to further improve the accuracy of the evaluation results. The reference samples are detailed in Table 8.

[0129] Table 8: Longjing tea multi-dimensional characteristic aroma attribute reference sample table

[0130]

[0131]

[0132] Ten experts conducted multi-dimensional quantitative description analysis on 18 aroma descriptors of three samples. The multi-dimensional characteristic aroma attribute scores are shown in Table 9.

[0133] Table 9: Quantified Si values ​​of the multi-dimensional characteristic aroma attributes of Longjing tea (Appendix Figure 3-5 )

[0134]

[0135] 6) Combining the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea after weighting in step 4) and the quantitative value of the intensity of the multi-dimensional characteristic aroma attribute of tea in step 5), the weighted total score of the aroma characteristics of the sample is calculated to form a comprehensive quantitative evaluation model of the aroma quality of tea.

[0136] Based on the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea obtained in step 4) and the quantitative value of the intensity of the multi-dimensional characteristic aroma attribute of tea obtained in step 5), a comprehensive quantitative evaluation model of tea aroma quality is formed, and the final score K=∑(M i ×S i ), as shown in Table 9.

[0137] Table 9: Weighted total score of aroma characteristics

[0138] Production area West Lake Qiantang Yuezhou Total score 3.531 3.558 2.180

[0139] Specifically, in this embodiment, the scoring results of the above-mentioned quantitative descriptive analysis method and the weights of the characteristic aroma attributes can be used to conclude that the final scores of the three Longjing tea samples from the West Lake production area, Qiantang production area, and Yuezhou production area are: 3.531 points for the West Lake production area; 3.558 points for the Qiantang production area; and 2.180 points for the Yuezhou production area, ultimately forming a comprehensive quantitative evaluation model for the aroma quality of Longjing tea.

[0140] In summary, this patent establishes a method for evaluating the multi-dimensional characteristic aroma of tea based on hierarchical analysis. For the first time, the aroma of tea leaves, the aroma of the soup surface, and the aroma after the nose are combined to quantitatively evaluate its multi-dimensional characteristic aroma attributes. Taking Longjing tea samples as an example, the present invention determines the multi-dimensional characteristic aroma of Longjing tea, and then combines the hierarchical analysis method to calculate the weight of the multi-dimensional characteristic aroma attributes, and uses the quantitative descriptive analysis method to score its intensity. Finally, combined with the preference effect of different characteristic aroma attributes, the weighted total score of the aroma characteristics of Longjing tea samples from three different production areas is calculated, and finally a comprehensive quantitative evaluation model for the aroma quality of Longjing tea is formed.

[0141] Comparative Example 1:

[0142] The difference between this embodiment and embodiment 1 is that this embodiment adopts a traditional sensory evaluation method, the specific contents of which are as follows:

[0143] The present invention is specifically introduced as follows using Zhejiang Longjing tea as a comparative example:

[0144] Tea samples were prepared and evaluated according to GB / T 23776-2018: The tea preparer weighed 3.0 g of Longjing tea from three different regions into a standard evaluation cup, poured 150 mL of boiling water, and brewed for 4 minutes. After filtering the tea, the aroma of the tea leaves in the evaluation cup was evaluated by one expert grader, and the tea leaves were retained for evaluation. The results are shown in Table 10.

[0145] Table 10 Sensory evaluation results of Longjing tea samples (GB / T 23776-2018)

[0146] Tea sample producing area Comments Score West Lake Noble, rich, long-lasting 94 Qiantang Fresh, rich and long-lasting 95 Yuezhou High and refreshing, tender and fragrant 92

[0147] In summary:

[0148] Comparison of Example 1 with Comparative Example 1: First, the GB / T 23776-2018 Tea Sensory Evaluation Method only evaluates the single dimension of leaf aroma within the sensory layer. Compared to the national standard GB / T 23776-2018, the method proposed in this patent expands the breadth of sensory layer evaluation, evaluating three dimensions: leaf aroma, tea aroma, and postnasal aroma.

[0149] Second, the qualitative results of quality given by the national standard method are comprehensive descriptions (such as: noble, rich, and lasting), while the qualitative results of this patent are 18 specific attributes of three perception layers (overall intensity, green smell, fresh fragrance, fresh bamboo shoot smell, winter leaf smell, tender fragrance, chestnut fragrance, fresh floral fragrance, sweet floral fragrance, honey fragrance, milk fragrance, sweet fragrance, fresh fragrance, fire, smoke, burnt air, stuffy air, and oily breath), which improves the accuracy of the perceived object.

[0150] Third, the quantitative result given by the national standard method is a comprehensive aroma score, with an evaluation scale of 100 points. According to Table 10, the sample score difference is 1 to 3 points. At this time, the quantitative difference in aroma scores between samples is 1%-3%.

[0151] The patent's quantitative results first quantify a single attribute, using a scale of 9. For example, for the floral aroma attribute of the base tea, the scores of the three tea samples ranged from 3.17 to 7.19, with a quantitative difference of 35.2% to 79.9%. Therefore, compared to the national standard method, the patent's evaluation results have a more sensitive perceptual quantitative difference.

[0152] Then, the hierarchical analysis method and the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea after empowering the preference effect are used. The evaluation results of this patent can characterize the aroma differences between different tea samples more scientifically and accurately from the perspective of multi-dimensional perception quantification.

Claims

1. A multi-dimensional tea aroma quantitative evaluation method based on hierarchical analysis, characterized in that: The following steps are involved: 1) Collect and prepare tea samples according to standards, and determine the multi-dimensional characteristic aroma attributes of tea based on the Chinese tea sensory aroma wheel; 2) constructing a hierarchical structure model of the multi-dimensional characteristic aroma attributes of tea consisting of multiple levels according to the multi-dimensional characteristic aroma attributes of tea obtained in step 1); 3) Based on the hierarchical analysis method, the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea obtained in step 2) is used to calculate the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea 3.1) Construct judgment matrices for different levels in the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea, and have expert evaluators score the importance of each sub-indicator of each level in the hierarchical structure model of the multi-dimensional characteristic aroma attribute of tea obtained in step 2), and further normalize the judgment matrix; 3.2) Based on the hierarchical analysis method, calculate the weight coefficients of the sub-indicators at each level of the hierarchical structure model of the multi-dimensional characteristic aroma attributes of the leaves in step 2), and then calculate the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of the tea leaves using the weight coefficients of the sub-indicators at each level; 3.3) Based on the consistency matrix method, the consistency of the importance scores obtained by the expert evaluators was evaluated; 4) Weighting the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea obtained in step 3) by preference effect, and further calculating the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea 4.1) Using the multi-dimensional characteristic aroma attributes of tea in step 1), analyze consumers' preferences for different attributes of tea to obtain the preference effect of each attribute; 4.2) Based on the preference effect, the preference effect is assigned to the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of the tea obtained in step 3), and then the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of the tea is calculated based on the weighted result; V) Based on the quantitative descriptive analysis method, expert evaluators quantitatively evaluated the intensity of the characteristic aroma attributes of tea samples in three dimensions: leaf aroma, soup aroma, and after-nasal aroma, and obtained the quantitative value of the intensity of the multi-dimensional characteristic aroma attributes of tea; 6) Combining the comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea after weighting in step 4) and the quantitative value of the intensity of the multi-dimensional characteristic aroma attribute of tea in step 5), the weighted total score of the aroma characteristics of the sample is calculated to form a comprehensive quantitative evaluation model of the aroma quality of tea.

2. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 1, characterized in that: In step 1), the specific contents of preparing tea samples according to the standard are as follows: 1.1) Weigh 3.0g of the collected tea leaves into a standard evaluation cup, add 150mL of boiling water, and brew for 4 minutes; 1.2) Separate the brewed tea sample, leave the tea leaves in a standard evaluation cup for testing, and pour the filtered tea into a standard evaluation bowl for testing; 1.3) Organize more than 10 expert evaluators, with an even number of expert evaluators, and divide all expert evaluators into multiple evaluation groups, each of which contains two expert evaluators; 1.4) Randomly code the standard evaluation cup containing tea leaves and the standard evaluation bowl containing tea soup. Present each pair of randomly coded standard evaluation cups and standard evaluation bowls to two expert evaluators from the same group. The two expert evaluators will evaluate the aroma of the tea leaves in the standard evaluation cups and the aroma and after-nose aroma of the tea soup in the standard evaluation bowls. The other evaluation groups will follow the same procedure. 1.5) Prepare the same tea sample according to 1.1) and 1.2) again. According to 1.4), present a pair of randomly coded standard evaluation cups and standard evaluation bowls to two expert evaluators in the same group. The two expert evaluators exchange the evaluation objects for aroma evaluation.

3. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 2, characterized in that: In step 1), the method for determining the multi-dimensional characteristic aroma attributes of tea is: 1.6) After evaluation by expert evaluators, the consensus and non-consensus attributes of the multi-dimensional characteristic aroma of the sample are confirmed. Non-consensus attributes should be repeatedly evaluated and opinions solicited through multiple rounds of samples until all attributes are consensus attributes and the final evaluation results are obtained; Among them, consensus attributes are attributes on which expert evaluators directly reach consensus, and non-consensus attributes are attributes on which no consensus is directly reached: 1.7) Generate a list of multi-dimensional characteristic aroma attributes of tea based on the final evaluation results of the expert evaluators.

4. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 1, characterized in that: The specific content of the step 2) is: the layers are divided into a core layer, a perception layer, a classification layer and a feature layer.

5. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 4, characterized in that: The content of step 3.1 is specifically as follows: 3.1.1) By comparing the importance of each sub-indicator at each level, refer to the following table to determine the score of the comparison result of the importance of sub-indicator i and sub-indicator j as a ij , based on the score a of the comparison results of the importance between indicators ij Generate judgment matrix A ij ,in, i and j represent two different sub-indicators, i, j = 1, 2, ..., n, and the importance comparison results of sub-indicators at each level are a. ij The reference standards are as follows: 3.1.2) For the judgment matrix A ij Perform normalization processing to obtain a normalized judgment matrix) in, n is the matrix order.

6. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 5, characterized in that: The specific content of step 3.2) is: 3.2.1) Calculate the weight coefficients of the sub-indicators at each level: Add row by row to get the eigenvector Then the eigenvector Perform normalization to obtain the index A in the single ranking of each level i Weight coefficient N for the superior indicator i , 3.2.2) Calculation of the comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea: Comprehensive weight coefficient of tea's multi-dimensional aroma attributes 7. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 6, wherein the specific contents of step 3.3) are: 3.3.1) For the maximum characteristic root λ of the judgment matrix max Perform the calculation: 3.3.2) Further, the matrix consistency index CI is calculated: 3.3.3) Further, the consistency ratio value CR is calculated: Among them, RI is the random consistency index. If CR is less than 0.1, the matrix is ​​considered to have good consistency. It is used to judge whether the consistency of the scoring results among expert evaluators is good. If it is good, the obtained weight coefficient is reliable.

8. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 7, wherein the step (iv) is as follows: The specific contents of step 4.1) are: 4.1.1) Based on the multi-dimensional characteristic aroma attributes of tea in step 1), survey consumers on their preferences for different attributes, evaluate the survey results, and obtain the preference effects of specific attributes at the characteristic level; 4.1.2) The specific criteria for distinguishing preference effects are as follows: a positive effect is defined as a percentage of consumers liking an attribute in the multi-dimensional aroma attribute exceeding 70%, a negative effect is defined as a percentage of consumers liking an attribute below 30%, and a neutral effect is defined as a percentage of consumers liking an attribute between 30% and 70%. The specific contents of step 4.2) are: M i =P i ×w, where M i is the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attributes of tea, where the weighting coefficient w of the positive effect characteristic layer sub-indicator is 1, the weighting coefficient w of the negative effect characteristic layer sub-indicator is -1, and the weighting coefficient w of the neutral effect characteristic layer sub-indicator is 0.

9. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 8, characterized in that: In the step 5), a multidimensional aroma evaluation method is proposed based on the quantitative description analysis method using the multidimensional characteristic aroma attributes obtained in the step 1), which specifically comprises: establishing a reference sample of the multidimensional characteristic aroma attributes, preparing and presenting tea samples, and organizing more than 10 expert evaluators to sequentially evaluate the specific attributes of the leaf bottom aroma, soup surface aroma, and after-nasal aroma to obtain the quantitative value S of the intensity of the multidimensional characteristic aroma attributes of the tea. i .

10. The method for quantitatively evaluating the multi-dimensional aroma of tea based on hierarchical analysis according to claim 9, characterized in that: In step 6), based on the weighted comprehensive weight coefficient of the multi-dimensional characteristic aroma attribute of tea obtained in step 4) and the quantitative value of the intensity of the multi-dimensional characteristic aroma attribute of tea obtained in step 5), a comprehensive quantitative evaluation model of the aroma quality of tea is formed, and the final score K=∑(M i ×S i ).

Images