A method for grading evaluation of spice quality based on probability calculation

Abstract

The application provides a kind of grading evaluation method of spice quality based on probability calculation, which comprises the following steps: 1) selection of spice determination basic index and data collection and statistics; 2) selection of characteristic quality index; 3) determination of characteristic index weight; 4) determination of quality level and probability; 5) determination of quality grading critical point; 6) development of quality grading standard of spice according to grading critical point. The application is a relatively accurate quality grading method based on quantitative index, which can to some extent get rid of quality acceptance method based on sense and experience, improve the objectivity of spice quality evaluation, and reduce the difficulty of quality evaluation.

Description

Technical Field

[0001] This invention belongs to the field of agricultural product processing technology, specifically a method for grading and evaluating the quality of spices. Background Technology

[0002] Spices come in a wide variety and are widely used in food processing and cooking, having a decisive influence on the flavor of food. Therefore, selecting high-quality spices is of great significance for improving the quality of the final food products.

[0003] In spice quality management, many quality standards typically combine sensory evaluation and appearance assessment. These methods are relatively general and, in practice, rely heavily on the experience of the inspectors, making precision and quantification difficult. Furthermore, some spices lack grading standards, leading to inconsistent spice quality in the market.

[0004] Therefore, there is an urgent need for professional, efficient, and convenient grading and evaluation methods to refine grading standards and achieve quality control of spices. Summary of the Invention

[0005] The purpose of this invention is to provide a method for grading and evaluating the quality of spices. Based on the characteristics of spices, this method combines traditional experience with modern methods to develop a scientific and systematic quality grading and evaluation method. It can accurately, conveniently, and effectively reflect the quality of spices, reduce the difficulty of spice quality evaluation, and achieve the goals of regulating the market, improving quality, and guiding a positive market trend of "high quality, high price."

[0006] To achieve the above-mentioned objectives, the specific technical solution of this invention is as follows:

[0007] A method for grading and evaluating the quality of spices includes the following steps:

[0008] 1) Selection of basic indicators for spice testing and data collection and statistics;

[0009] 2) Selection of characteristic quality indicators;

[0010] 3) Determining the weights of feature indicators;

[0011] 4) Determining the quality level and probability;

[0012] 5) Determination of the critical point for quality grading;

[0013] 6) Establish quality grading standards for spices based on grading thresholds.

[0014] Furthermore, the specific steps for collecting and statistically analyzing spice quality-related data in step 1) are as follows:

[0015] Multiple samples of a representative spice were collected through various channels. Combining the spice's own characteristics with relevant literature information, data on spice quality-related indicators were comprehensively measured to obtain a quality database. The distribution range of each indicator was determined and converted into data that conforms to a normal distribution.

[0016] Table of quality-related indicators for common spice categories:

[0017] Spice Quality Index Table

[0018]

[0019] In the table above, all indicators are characterized as "continuously distributed, quantifiable, and capable of having a definite positive or negative impact on quality"; among the "other" indicators, some indicators have results close to 100% in actual spice tests and are not applicable to all spices.

[0020] The definitions of some indicators in the table of quality-related indicators for common spice categories are as follows:

[0021] (6) Fruit weight: The mass of a unit quantity of fruit, seeds, and flowers;

[0022] (7) Integrity rate: The percentage of spice samples per unit mass that retains the original spice integrity.

[0023] (8) Defect rate: The percentage of defective samples per unit mass of spice sample due to incomplete development, browning, mold or insect infestation.

[0024] (9) Non-volatile ether extract: a liquid substance with certain chemical components extracted from spices;

[0025] (10) Impurity rate: The percentage of substances other than spices in the product.

[0026] Furthermore, the specific steps for selecting characteristic quality indicators in step 2) are as follows:

[0027] (5) Based on the part of the spice used, refer to the table of quality-related indicators of common spice categories and measure the corresponding indicators.

[0028] (6) Based on the function of the selected spice, select one indicator from each function in the table of quality-related indicators of common spice categories for measurement;

[0029] (7) Based on the characteristics of the selected spices, determine some of the "Other" indicators in the table of quality-related indicators for common spice categories;

[0030] (8) Based on the correlation analysis results of all the above-mentioned measurement index data, the final grading index is determined to ensure that there is no significant relationship between the retained quality indexes.

[0031] Furthermore, the specific steps for weighting the feature indicators in step 3) are as follows:

[0032] 1) Importance score

[0033] Using the expert survey method within the subjective weighting approach, experts' knowledge and experience were pooled to score the spices based on the importance of each indicator. The importance score table is as follows:

[0034] Spices Importance Score Evaluation Table

[0035]

[0036] (1) Selection of experts: By visiting spice and condiment processing enterprises, 10 to 30 experts with both practical work experience and deep theoretical knowledge were selected to participate in the determination of weights.

[0037] (2) Initial expert evaluation: The n indicators with undetermined weights, relevant data, and unified rules for determining weights are sent to the selected experts, who are asked to independently give the importance score of each indicator.

[0038] (3) Collect the results and calculate the mean and standard deviation of the importance scores for each indicator;

[0039] (4) Return the calculation results and supplementary information to the experts and ask them to determine the scores based on the new information.

[0040] (5) Repeat steps (3) and (4). When giving the final weight, the experts need to indicate the confidence level of their respective scores in order to make the judgment more accurate until the difference between the importance score of each indicator and its mean does not exceed the pre-given standard, that is, the opinions of the experts are basically consistent.

[0041] 2) Coefficient of variation α b Calculation

[0042] The formulas for calculating the coefficient of variation of various spice indicators are as follows;

[0043]

[0044] In the above formula, S k A represents the standard deviation of the k-th indicator. k This represents the mean of the k-th indicator.

[0045] 3) Calculation of weights (1) The importance scores or coefficients of variation of different indicators are denoted as X1, X2, ..., X nThen the weight a of the k-th indicator k (Keep 2 decimal places) Calculate as shown in formula (2):

[0046]

[0047] In the above formula, X k Let be the score or coefficient of variation for the k-th indicator. The weight calculated from the coefficient of variation is denoted as the objective weight 'a'. cv The weight calculated by expert scoring is denoted as subjective weight a. s ;

[0048] (2) The final weight 'a' of each indicator is determined jointly by the objective weight and the subjective weight (the calculation result is rounded to three decimal places). The calculation formula is as follows:

[0049] α=α cv *0.3+a s *0.7……………………………(3)

[0050] Furthermore, the specific steps for determining the quality level and probability in step 4) are as follows: Based on the actual situation, the spices are divided into several grades according to their quality, and each grade is numbered in order from best to worst. The best grade is recorded as grade 1, and so on.

[0051] The sample classification shall be determined according to the following criteria:

[0052] (1) If all quality indicators involved in the grade classification reach or exceed a certain pre-defined critical point, then the sample meets the requirements for classification into that grade.

[0053] (2) If at least one of the quality indicators involved in the grading fails to reach a predetermined critical point, the sample cannot be classified into that grade, but should be classified into the grade of the worst indicator.

[0054] Here, the probability (V) of a certain level is defined as the probability that a randomly selected spice sample will be in a certain specific level.

[0055] Define the cumulative probability (P) of level n. n For a given sample, the probability that all quality indicators involved in the grade classification reach or exceed the critical point specified by that indicator is;

[0056] Define the probability V of each level occurring under random conditions, and the cumulative probability P of level n. n The calculation is as shown in formula (4):

[0057]

[0058] In the formula, V iLet i be the probability of occurrence of level i.

[0059] Furthermore, the specific steps for determining the critical point of quality grading in step 5) are as follows: based on the probability corresponding to the quality level, the critical point of each indicator grading is determined by weighting coefficients.

[0060] 5-1) Determination of the probability of grade classification for each evaluation index of spices

[0061] Based on the overall evaluation index grading probability values ​​of existing spice samples, the grading probability requirements for each evaluation index are determined. For a quality index k, if a smaller observed value indicates better quality, then the cumulative probability R of the spice characteristic quality index k reaching the grade n requirement is determined. kn (%) is calculated using formula (5):

[0062]

[0063] For a quality index k, if a larger observed value indicates better quality, then the cumulative probability R of the spice characteristic quality index k reaching grade n is... kn (%) is calculated using formula (4):

[0064]

[0065] In the formula: R kn —The cumulative probability (%) of the characteristic quality index k of spices meeting the requirements of grade n;

[0066] P n —The cumulative probability (%) of the nth level of spices;

[0067] a k —The weight of the spice index k;

[0068] 5-2) Determination of the critical points for each indicator classification

[0069] Based on the cumulative probability of each indicator calculated above, and combined with the normal distribution curve of the indicator, we back-calculate its value under the cumulative probability, which serves as the critical value for grading.

[0070] Further, in step 6), a grading standard is developed based on the calculated critical points, merging those parts whose critical points are too close to be measurable.

[0071] Compared with the prior art, the positive effects of the present invention are reflected in:

[0072] Based on the characteristics of spices, this invention combines traditional experience with modern methods to develop a scientific and systematic quality grading and evaluation method. This method can accurately, conveniently and effectively reflect the quality of spices, reduce the difficulty of spice quality evaluation, and achieve the goals of standardizing the market, improving quality, and guiding a positive market trend of high quality and high price. Attached Figure Description

[0073] Figure 1 This is a flowchart illustrating the spice quality grading and evaluation method described in this invention. Detailed Implementation

[0074] To facilitate understanding of the present invention, specific embodiments are described below. However, this should not be construed as limiting the scope of the invention to the following embodiments.

[0075] In the following embodiments, the spice quality index table is shown in Table 1:

[0076] Table 1. Quality Indicators of Spices

[0077]

[0078]

[0079] The importance score table is shown in Table 2:

[0080] Table 2. Importance Score Evaluation Table for Spices

[0081]

[0082] Example 1:

[0083] The quality grading and evaluation method for Amomum villosum includes the following steps:

[0084] 1) Collection and statistics of quality-related data for the spice Amomum villosum;

[0085] 1-1 Sample Collection

[0086] We collected 100 samples of the same spice from multiple channels, including wholesale markets, supermarkets, Taobao, and condiment companies.

[0087] 1-2 Selection and Determination of Quality-Related Indicators for Amomum villosum

[0088] Based on Table 1, the final selected test indicators are: fruit weight, size (length, width), impurity rate, volatile oil, moisture, total ash, defective product rate, and intact product rate.

[0089] 1-3 Data Preprocessing

[0090] (1) Summarize the observed values ​​of each indicator to ensure the accuracy of the data;

[0091] (2) For all the observed values ​​of a certain indicator, check whether each indicator value conforms to a normal distribution through observation or mathematical verification. If the data of the indicator shows a normal distribution, subsequent judgment can be made. Otherwise, the data is mathematically transformed to make it show a normal distribution for convenient unified analysis.

[0092] 1-4 Determination of the mean and standard deviation of normally distributed data

[0093] The distribution of various indicators of Amomum villosum is as follows, based on calculations:

[0094] (1) The observed values ​​of the index “defect rate” are taken as the logarithm with base e, with a mean of 1.41 and a standard deviation of 0.950. It can be approximated as a normal distribution and is denoted as “Ln defect rate ~ N(1.41, 0.902)”.

[0095] (2) The indicator “complete product rate” can be approximated as a normal distribution with a mean of 86.91 and a standard deviation of 5.623, denoted as “complete product rate ~ N(86.91, 31.617)”;

[0096] (3) The observed values ​​of the index “impurity rate” are taken as the logarithm to the base e, with a mean of -0.84 and a standard deviation of 1.056. They can be approximated as a normal distribution and are denoted as “Ln impurity rate ~ N(-0.84, 1.116)”.

[0097] (4) The index “length” has a mean of 2.17 and a standard deviation of 0.073, which can be approximated as a normal distribution, denoted as “length ~ N(2.17, 0.005)”;

[0098] (5) The index “width” can be approximated as a normal distribution with a mean of 1.51 and a standard deviation of 0.093, denoted as “short ~ N(1.51, 0.093)”;

[0099] (6) The indicator “moisture” can be approximated as a normal distribution with a mean of 8.63 and a standard deviation of 1.293, denoted as “moisture ~ N(8.63, 1.673)”;

[0100] (7) The indicator “total ash content” can be approximated as a normal distribution with a mean of 5.90 and a standard deviation of 0.912, denoted as “total ash content ~ N(5.90, 0.832)”;

[0101] (8) The index “volatile oil” can be approximated as a normal distribution with a mean of 1.11 and a standard deviation of 0.264, denoted as “volatile oil ~ N(1.11, 0.070)”;

[0102] (9) The indicator “fruit weight” can be approximated as a normal distribution with a mean of 97.90 and a standard deviation of 14.485, denoted as “fruit weight ~ N(97.90, 209.828)”;

[0103] 2) Selection of quality indicators for Amomum villosum;

[0104] (1) For Amomum villosum, the content of volatile oil is a key indicator determining its flavor quality. Therefore, using volatile oil as the key indicator, a correlation analysis was performed on all measured indicators, and the results are as follows:

[0105] Table 3. Correlation P-values ​​of various indicators of Amomum villosum

[0106]

[0107]

[0108] In the table above, a p-value of no more than 0.05 indicates that the two indicators are considered to be correlated.

[0109] 1) Besides volatile oil, the defect rate, impurity rate, intact rate, moisture content, total ash content, and fruit weight can all affect the appearance and shelf life of Amomum villosum. The correlation between dimensional indicators and the quality of Amomum villosum is unclear; therefore, the length and width indicators are removed.

[0110] 2) There was no significant correlation between volatile oil and six indicators: defect rate, impurity rate, integrity rate, moisture content, total ash content, and fruit weight. Subsequent studies will only retain seven indicators: volatile oil, defect rate, impurity rate, integrity rate, moisture content, total ash content, and fruit weight.

[0111] 3) Determination of the weights of characteristic indicators of Amomum villosum;

[0112] 3-1 The coefficient of variation of the seven indicators was calculated based on the observation data and formula (1). The results are shown in Table 1:

[0113] Table 4. Coefficients of variation for seven quality indicators of Amomum villosum.

[0114] volatile oil 0.24 Defect rate 0.95 Moisture 0.15 Complete product rate 0.06 Total Ash 0.15 heavy fruit 0.15 Impurity rate 1.01

[0115] According to formula (2), the objective weight 'a' of each indicator is calculated based on the coefficient of variation. cv The results are shown in Table 5:

[0116]

[0117] In the above formula, X1, X2...X7 are the coefficients of variation for the seven indicators: volatile oil, moisture, total ash, impurity rate, defective product rate, intact product rate, and fruit weight. k Let a be the coefficient of variation of the k-th indicator. cv The weights obtained from the calculation of the coefficient of variation are denoted as objective weights.

[0118] Table 5 Objective weights of seven quality indicators of Amomum villosum a cv

[0119] volatile oil 0.09 Defect rate 0.35 Moisture 0.06 Complete product rate 0.02 Total Ash 0.06 heavy fruit 0.05 Impurity rate 0.37

[0120] 3-2 By visiting spice and seasoning processing enterprises and using expert research methods, corresponding scores were assigned according to the importance of each quality indicator of Amomum villosum:

[0121] Table 6 Importance Scores of Seven Quality Indicators for Amomum villosum

[0122] volatile oil 7 Defect rate 4 Moisture 4 Complete product rate 3 Total Ash 2 heavy fruit 3 Impurity rate 6

[0123] According to formula (2), the subjective weight a of each indicator is calculated based on the coefficient of variation. s The results are shown in Table 7:

[0124]

[0125] In the above formula, X1, X2...X7 are the scores for the seven indicators: volatile oil, moisture, total ash, impurity rate, defective product rate, intact product rate, and fruit weight, respectively. k For the score of the k-th indicator, a s The weights obtained from the importance scores are denoted as subjective weights.

[0126] Table 7 Subjective weights of seven quality indicators of Amomum villosum a s

[0127] volatile oil 0.24 Defect rate 0.14 Moisture 0.07 Complete product rate 0.10 Total Ash 0.14 heavy fruit 0.10 Impurity rate 0.21

[0128] 3-3 Summing is performed after weighting according to objective and subjective weights respectively (a) cv ×0.3+a s (×0.7) Determine the final weight 'a' for each indicator:

[0129] Table 8. Final weights of seven quality indicators for Amomum villosum (a)

[0130]

[0131]

[0132] 4) Determination of the quality grade and probability of Amomum villosum;

[0133] 4-1 Quality Level Classification

[0134] Based on the needs, Sichuan Amomum villosum is subjectively classified into grades 1, 2, 3, 4, and 5. The lower the grade number, the higher the quality.

[0135] 4-2 Determining the probability of occurrence of each quality level

[0136] As needed, the probability of each grade of Amomum villosum appearing is subjectively determined, and the cumulative probability of each grade is calculated according to formula (4):

[0137] Table 9. Probability (V) and cumulative probability (P) of the five quality grades of Sichuan Amomum villosum.

[0138] 1 0.5％ 0.5％ 2 4.5％ 5％ 3 45％ 50％ 4 40％ 90％ 5 10％ 100％

[0139] 5) Based on the probability corresponding to the quality level, determine the critical point for grading each indicator of Amomum villosum using weighting coefficients;

[0140] 5-1 Determination of the probability (R) of each evaluation indicator's level classification

[0141] Based on the overall evaluation index level classification probability value (P) and the weights (a) of each index of Amomum villosum mentioned above, the cumulative probability density value (R) corresponding to the evaluation index on its 50% normal curve when the requirements of each level are met is calculated. The calculation results are shown in Table 2:

[0142] Table 10 Cumulative probability density values ​​of various evaluation index grades for Amomum villosum

[0143]

[0144] 5-2 Determination of the critical points for each indicator classification

[0145] Based on the cumulative probability density determined above, the values ​​corresponding to the normal distribution of each index under different cumulative probability density conditions are calculated, which serve as the critical points for index classification.

[0146] Table 11 Critical points for various evaluation indicators of Amomum villosum

[0147]

[0148] 6) Compile grading standards based on the calculated critical points, and merge those with critical points that are too close to each other and difficult to measure;

[0149] Table 12 Grading Standards for Various Evaluation Indicators of Amomum villosum

[0150]

[0151] Example 2:

[0152] The method for grading and evaluating the quality of fennel seeds includes the following steps:

[0153] 1) Collection and statistics of quality-related data for the spice fennel;

[0154] 1-1 Sample Collection: 200 fennel samples were collected from multiple channels, including wholesale markets, supermarkets, Taobao, and condiment-related companies;

[0155] 1-2 Selection and determination of quality-related indicators for fennel;

[0156] Based on Table 1, the final selected test indicators are: fruit weight, size (length, width), impurity rate, volatile oil, moisture, total ash, and defective product rate.

[0157] 1-3 Data Preprocessing

[0158] (1) Summarize the observed values ​​of each indicator to ensure the accuracy of the data;

[0159] (2) For all observed values ​​of a certain indicator, check whether each indicator value conforms to a normal distribution through observation or mathematical verification. If the data of the indicator shows a normal distribution, subsequent judgment can be made. Otherwise, the data is mathematically transformed to make it show a normal distribution for convenient unified analysis.

[0160] 1-4 Determination of the mean and standard deviation of normally distributed data

[0161] The distribution of various indicators of fennel seeds, based on calculations, is as follows:

[0162] (1) The index “volatile oil” can be approximated as a normal distribution with a mean of 2.19 and a standard deviation of 0.484, denoted as “volatile oil ~ N(2.19, 0.234)”;

[0163] (2) The indicator “total ash content” can be approximated as a normal distribution with a mean of 8.14 and a standard deviation of 0.409, denoted as “total ash content ~ N(8.14, 0.167)”;

[0164] (3) The indicator “moisture” can be approximated as a normal distribution with a mean of 6.75 and a standard deviation of 1.357, denoted as “moisture ~ N(6.75, 1.842)”;

[0165] (4) The index “impurity rate” can be approximated as a normal distribution with a mean of 3.07 and a standard deviation of 1.455, denoted as “impurity rate ~ N(3.07, 1.455)”.

[0166] (5) The index “length” can be approximated as a normal distribution with a mean of 0.68 and a standard deviation of 0.044, denoted as “length ~ N(0.68, 0.002)”;

[0167] (6) The index “width” can be approximated as a normal distribution with a mean of 0.21 and a standard deviation of 0.020, denoted as “short ~ N(0.21, 0.000)”;

[0168] (7) The indicator “fruit weight” can be approximated as a normal distribution with a mean of 0.59 and a standard deviation of 0.085, denoted as “fruit weight ~ N(0.59, 0.085)”;

[0169] (8) The index “defect rate” can be approximated as a normal distribution with a mean of 0.53 and a standard deviation of 0.415, and is denoted as “defect rate ~ N(0.53, 0.415)”.

[0170] 2) Selection of representative quality indicators for fennel;

[0171] (1) For fennel, the content of volatile oil is the key indicator that determines its flavor quality; therefore, using the volatile oil content as the key indicator, a correlation analysis was performed on all the measured indicators, and the results are as follows:

[0172] Table 13 Correlation P-values ​​of Various Indicators of Fennel

[0173]

[0174] In the table above, a p-value of no more than 0.05 indicates that the two indicators are considered to be correlated.

[0175] 1) Besides volatile oil, total ash, moisture, impurity rate, number of fruits, and defective product rate can all affect the appearance and shelf life of fennel. The correlation between size-based indicators and the quality of fennel is unclear; therefore, length and width indicators are removed.

[0176] 2) There was no significant correlation between volatile oil and the five indicators of impurity rate, fruit weight, moisture, total ash, and defect rate. Subsequent studies only retained six indicators: volatile oil, total ash, moisture, impurity rate, defect rate, and fruit weight.

[0177] 3) Determination of the weights of fennel characteristic indicators;

[0178] 3-1 Calculate the coefficient of variation based on the observed data of the six indicators:

[0179] Table 14 Coefficients of Variation for Six Quality Indicators of Fennel

[0180] volatile oil 0.22 Impurity rate 0.47 Total Ash 0.05 heavy fruit 0.14 Moisture 0.20 Defect rate 0.79

[0181] According to formula (2), the objective weight 'a' of each indicator is calculated based on the coefficient of variation. cv The results are shown in Table 5:

[0182]

[0183] In the above formula, X1, X2...X6 are the coefficients of variation of the six indicators: volatile oil, total ash, moisture, impurity rate, defective product rate, and fruit weight. k Let a be the coefficient of variation of the k-th indicator. cv The weights obtained by calculating the coefficient of variation are denoted as objective weights.

[0184] Table 15 Objective weights of six quality indicators for fennel seeds (a)cv

[0185] volatile oil 0.12 Impurity rate 0.25 Total Ash 0.03 heavy fruit 0.08 Moisture 0.11 Defect rate 0.42

[0186] 3-2 By visiting spice and seasoning processing enterprises and using the expert survey method, corresponding scores were assigned to the importance of each quality indicator of fennel in an importance scoring table. The scoring opinions of 15 experts were collected and calculated. The final results are shown in Table 6:

[0187] Table 16 Importance Scores of Six Quality Indicators for Fennel Seeds

[0188] volatile oil 8 Impurity rate 3 Total Ash 3 heavy fruit 3 Moisture 5 Defect rate 4

[0189] According to formula (2), the subjective weight a of each indicator is calculated based on the coefficient of variation. s The results are shown in Table 17:

[0190]

[0191] In the above formula, X1, X2...X6 represent "volatile oil, total ash, moisture, impurity rate, defective product rate, and fruit weight," respectively. k For the score of the k-th indicator, a s The weights obtained from the importance scores are denoted as subjective weights.

[0192] Table 17 Subjective weights of six quality indicators for fennel seeds (a) s

[0193] volatile oil 0.31 Impurity rate 0.12 Total Ash 0.12 heavy fruit 0.12 Moisture 0.19 Defect rate 0.15

[0194] 3-3 Summing is performed after weighting according to objective and subjective weights respectively (a) cv ×0.3+a s (×0.7) Determine the final weight 'a' for each indicator:

[0195] Table 18 Final weights of six quality indicators for fennel seeds (a)

[0196] volatile oil 0.251 Impurity rate 0.157 Total Ash 0.089 heavy fruit 0.104 Moisture 0.167 Defect rate 0.234

[0197] 4) Determination of the quality grade and probability of fennel seeds;

[0198] 4-1 Quality Level Classification

[0199] Based on the needs, fennel seeds are subjectively classified into grades 1, 2, 3, and 4, with the lower the grade number indicating higher quality.

[0200] 4-2 Determining the probability of occurrence of each quality level

[0201] As needed, the probability of each grade of fennel is subjectively determined, and the cumulative probability of each grade is calculated according to formula (4):

[0202] Table 19. Probability (V) and cumulative probability (P) of the four quality grades of fennel.

[0203] 1 1％ 1％ 2 19％ 20％ 3 70％ 90％ 4 10％ 100％

[0204] 5) Based on the probability corresponding to the quality level, determine the critical point for grading each indicator of fennel by using weighting coefficients;

[0205] 5-1 Determination of the probability (R) of each evaluation indicator's level classification

[0206] Based on the overall evaluation index level classification probability value (P) and the weights (a) of each index, the cumulative probability density value (R) corresponding to each evaluation index on its normal curve when meeting the requirements of each level is calculated. The calculation results are shown in Table 2:

[0207] Table 20 Cumulative probability density values ​​for the grade classification of fennel seeds by evaluation index

[0208]

[0209] 5-2 Determination of the critical points for each indicator classification

[0210] Based on the cumulative probability density determined above, the values ​​corresponding to the normal distribution of each indicator under different cumulative probability density conditions are calculated, which serve as the critical points for indicator classification.

[0211] Table 21 Critical points for various evaluation indicators of fennel

[0212]

[0213] 6) Compile grading standards based on the calculated critical points, and merge the parts whose critical points are too close to be measurable.

[0214] Table 22 Grading Standards for Various Evaluation Indicators of Fennel

[0215]

[0216] Example 3:

[0217] The method for grading and evaluating the quality of bay leaves includes the following steps:

[0218] 1) Collection and statistics of data related to the quality of spices and bay leaves;

[0219] 1-1 Sample Collection: 50 bay leaf samples were collected from multiple channels, including wholesale markets, supermarkets, Taobao, and condiment-related companies;

[0220] 1-2 Selection and determination of indicators related to bay leaf quality.

[0221] Based on Table 1, the final selected test indicators are: defective product rate, intact product rate, impurity rate, dimensions (length and width), volatile oil, moisture, and total ash content;

[0222] 1-3 Data Preprocessing

[0223] (1) Summarize the observed values ​​of each indicator to ensure the accuracy of the data;

[0224] (2) For all observed values ​​of a certain indicator, check whether each indicator value conforms to a normal distribution through observation or mathematical verification. If the data of the indicator shows a normal distribution, subsequent judgment can be made. Otherwise, the data is mathematically transformed to make it show a normal distribution for convenient unified analysis;

[0225] 1-4 Determination of the mean and standard deviation of normally distributed data

[0226] The distribution of various indicators for bay leaves was calculated as follows:

[0227] (1) The index “volatile oil” can be approximated as a normal distribution with a mean of 1.15 and a standard deviation of 0.355, denoted as “volatile oil ~ N(1.15, 0.216)”;

[0228] (2) The indicator “total ash content” can be approximated as a normal distribution with a mean of 4.06 and a standard deviation of 0.273, denoted as “total ash content ~ N(4.06, 0.074)”;

[0229] (3) The indicator “moisture” can be approximated as a normal distribution with a mean of 7.33 and a standard deviation of 0.528, denoted as “moisture ~ N(7.33, 0.279)”;

[0230] (4) The observed values ​​of the index “impurity rate” are logarithmic to the base e, with a mean of 0.75 and a standard deviation of 1.393. They can be approximated as a normal distribution and are denoted as “Ln impurity rate ~ N(0.75, 1.941)”.

[0231] (5) The index “defect rate” can be approximated as a normal distribution with a mean of 65.98 and a standard deviation of 12.605, denoted as “defect rate ~ N(65.98, 158.895)”;

[0232] (6) The indicator “complete product rate” can be approximated as a normal distribution with a mean of 43.27 and a standard deviation of 12.771, denoted as “complete product rate ~ N(43.27, 163.106)”;

[0233] (7) The index “length” can be approximated as a normal distribution with a mean of 6.66 and a standard deviation of 0.502, denoted as “length ~ N(6.66, 0.252)”;

[0234] (8) The index “width” can be approximated as a normal distribution with a mean of 2.83 and a standard deviation of 0.148, denoted as “width ~ N(2.83, 0.022)”;

[0235] 2) Selection of representative quality indicators for bay leaves;

[0236] (1) For bay leaves, the content of volatile oils is a key indicator determining their flavor quality. Therefore, using volatile oil content as the key indicator, a correlation analysis was performed on all measured indicators, and the results are as follows:

[0237] Table 23 Correlation P-values ​​of Various Indicators for Bay Leaves

[0238]

[0239]

[0240] In the table above, a p-value of no more than 0.05 indicates that the two indicators are considered to be correlated.

[0241] 1) Besides volatile oils, total ash, moisture, impurity rate, intact product rate, and defective product rate can all affect the appearance and shelf life of bay leaves. The correlation between dimensions (length and width) and the quality of bay leaves is unclear; therefore, the length and width indicators are removed.

[0242] 2) There was no significant correlation between volatile oil and the five indicators of total ash, moisture, impurity rate, intact product rate and defective product rate. Subsequent studies will only retain six indicators: volatile oil, total ash, moisture, impurity rate, intact product rate and defective product rate.

[0243] 3) Determining the weights of characteristic indicators for bay leaves;

[0244] 3-1 The coefficient of variation of the six indicators was calculated based on the observation data and formula (1), and the results are shown in Table 24:

[0245] Table 24. Coefficients of variation for six quality indicators of bay leaves

[0246] volatile oil 0.31 Impurity rate 1.03 Moisture 0.07 Complete product rate 0.30 Total Ash 0.07 Defect rate 0.19

[0247] According to formula (2), the objective weight 'a' of each indicator is calculated based on the coefficient of variation. cv The results are shown in Table 25:

[0248]

[0249] In the above formula, X1, X2...X6 are the coefficients of variation for the six indicators: volatile oil, moisture, total ash, impurity rate, intact product rate, and defective product rate. kLet a be the coefficient of variation of the k-th index. cv The weights obtained from the calculation of the coefficient of variation are denoted as objective weights.

[0250] Table 25 Objective weights of six quality indicators for bay leaves (a) cv

[0251] volatile oil 0.16 Impurity rate 0.52 Moisture 0.04 Complete product rate 0.15 Total Ash 0.03 Defect rate 0.10

[0252] 3-2 By visiting spice and condiment processing enterprises and using the expert survey method, corresponding scores were assigned to bay leaves according to the importance of each quality indicator in the importance scoring table. The scoring opinions of 16 experts were collected and calculated. The final results are shown in Table 6:

[0253] Table 26 Importance Scores of Six Quality Indicators for Bay Leaves

[0254] volatile oil 7 Impurity rate 3 Moisture 4 Complete product rate 3 Total Ash 2 Defect rate 3

[0255] According to formula (2), the subjective weight a of each indicator is calculated based on the coefficient of variation. s The results are shown in Table 27:

[0256]

[0257] In the above formula, X1, X2...X6 are the scores for the six indicators: volatile oil, moisture, total ash, impurity rate, intact product rate, and defective product rate, respectively. k For the score of the k-th indicator, a s The weights obtained from the importance scores are denoted as subjective weights.

[0258] Table 27 Subjective weights of six quality indicators for bay leaves (a) s

[0259] volatile oil 0.32 Impurity rate 0.14 Moisture 0.18 Complete product rate 0.14 Total Ash 0.09 Defect rate 0.14

[0260] 3-3 Summing is performed after weighting according to objective and subjective weights respectively (a) cv ×0.3+a s (×0.7) Determine the final weight 'a' for each indicator:

[0261] Table 28 Final weights of the six quality indicators for bay leaves (a)

[0262] volatile oil 0.270 Impurity rate 0.253 Moisture 0.138 Complete product rate 0.141 Total Ash 0.074 Defect rate 0.125

[0263] 4) Determination of the quality grade and probability of bay leaves;

[0264] 4-1 Quality Level Classification

[0265] Based on the needs, the quality of bay leaves is subjectively divided into four grades: Grade 1, Grade 2, Grade 3, and Grade 4. The lower the grade number, the higher the quality.

[0266] 4-2 Determining the probability of occurrence of each quality level

[0267] As needed, the probability of each grade of fennel is subjectively determined, and the cumulative probability of each grade is calculated according to formula (4):

[0268] Table 29. Probability (V) and cumulative probability (P) of the four quality grades of bay leaves.

[0269] 1 1％ 1％ 2 19％ 20％ 3 70％ 90％ 4 10％ 100％

[0270] 5) Based on the probability corresponding to the quality level, determine the critical point for grading each indicator of bay leaf by using weighting coefficients;

[0271] 5-1 Determination of the probability (R) of each evaluation indicator's level classification

[0272] Based on the probability values ​​(P) of the overall evaluation index of bay leaves and the weights (a) of each index, the cumulative probability density values ​​(R) of the evaluation index on its normal curve when the requirements of each level are met are calculated. The calculation results are shown in Table 30.

[0273] Table 30 Cumulative probability density values ​​for the grade classification of bay leaves by various evaluation indicators

[0274]

[0275] 5-2 Determination of the critical points for each indicator classification

[0276] Based on the cumulative probability density determined above, the values ​​corresponding to the normal distribution of each index under different cumulative probability density conditions are calculated, which serve as the critical points for index classification.

[0277] Table 31 Critical points for various evaluation indicators of bay leaves

[0278]

[0279] 6) Compile grading standards based on the calculated critical points, and merge those with critical points that are too close to each other and difficult to measure;

[0280] Table 32 Grading Standards for Various Evaluation Indicators of Bay Leaves

[0281]

[0282] Evaluation method verification:

[0283] Fifty samples of Sichuan cardamom, fennel, and bay leaves were collected from various channels including wholesale markets, supermarkets, Taobao, and condiment-related companies. Based on the aforementioned patent grading standards, grading verification was conducted, and the final grading results are shown in Tables 33, 34, and 35.

[0284] Table 33 Grading data of Amomum villosum samples

[0285] 1 3 11 3 21 4 31 3 41 3 2 2 12 3 22 4 32 4 42 3 3 3 13 4 23 4 33 1 43 4 4 4 14 4 24 3 34 3 44 4 5 3 15 3 25 3 35 4 45 4 6 4 16 2 26 3 36 3 46 4 7 4 17 2 27 4 37 3 47 3 8 4 18 3 28 4 38 4 48 5 9 3 19 3 29 4 39 4 49 5 10 3 20 3 30 5 40 4 50 3

[0286] Table 34. Grading data of fennel seeds.

[0287]

[0288]

[0289] Table 35 Grading Data of Bay Leaf Samples

[0290] 1 3 11 3 21 3 31 3 41 3 2 3 12 2 22 2 32 3 42 3 3 3 13 2 23 3 33 3 43 3 4 3 14 4 24 2 34 4 44 3 5 2 15 3 25 3 35 4 45 2 6 4 16 2 26 3 36 3 46 2 7 3 17 3 27 3 37 4 47 2 8 2 18 4 28 3 38 2 48 3 9 3 19 3 29 2 39 3 49 3 10 3 20 3 30 3 40 3 50 3

[0291] As shown in Table 33, among the 50 samples of Amomum villosum, the number of samples at grades 1, 2, 3, 4, and 5 were 1, 3, 22, 21, and 3, respectively, which basically conforms to the grading probability model of Amomum villosum. As shown in Table 34, among the 50 samples of Fennel, the number of samples at grades 1, 2, 3, and 4 were 1, 8, 37, and 4, respectively, which basically conforms to the grading probability model of Fennel. As shown in Table 35, among the 50 samples of Bay Leaf, the number of samples at grades 1, 2, 3, and 4 were 0, 12, 32, and 6, respectively, which basically conforms to the grading probability model of Bay Leaf. The above results further verify the accuracy and reliability of the grading method in this study.

[0292] The embodiments described above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. It should be noted that, for those skilled in the art, several improvements and modifications can be made based on the technical solution and patent concept of the present invention without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for grading and evaluating the quality of spices, characterized in that... Includes the following steps: 1) Selection of basic indicators for spice testing and data collection and statistics, the specific steps are as follows: Multiple samples of a representative spice were collected through various channels. Combining the spice's own characteristics with relevant literature information, data on spice quality-related indicators were comprehensively measured to obtain a quality database. The distribution range of each indicator was determined and converted into data that conforms to a normal distribution. The quality indicators of spices include: (1) Fruit weight: the mass of a unit quantity of fruit, seeds, and flowers; (2) Integrity rate: The percentage of spice samples per unit mass that retain the original complete form of the spice; (3) Defect rate: The percentage of defective samples per unit mass of spice sample due to incomplete development, browning, mold or insect infestation. (4) Non-volatile ether extract: a liquid substance with certain chemical components extracted from spices; (5) Impurity rate: The percentage of substances other than spices in the product; 2) Selection of characteristic quality indicators; 3) Determining the weights of feature indicators involves the following steps: 1) Importance score Using the expert survey method within the subjective weighting approach, experts' knowledge and experience were pooled to score the spices based on the importance of each indicator. The importance score table is shown below: (1) Selection of experts: By visiting spice and condiment processing enterprises, 10 to 30 experts with both practical work experience and deep theoretical knowledge were selected to participate in the determination of weights; (2) Initial evaluation by experts: The n indicators with undetermined weights, relevant data, and unified rules for determining weights are sent to the selected experts, who are asked to independently give the importance score of each indicator. (3) Collect the results and calculate the mean and standard deviation of the importance scores for each indicator; (4) Return the calculation results and supplementary information to all experts and ask them to determine the scores based on the new data; (5) Repeat steps (3) and (4). When giving the final weight, the experts need to indicate the confidence level of their respective scores in order to make the judgment more accurate until the difference between the importance score of each indicator and its mean does not exceed the pre-given standard, that is, the opinions of the experts are basically consistent. 2) Calculation of coefficient of variation The formulas for calculating the coefficient of variation of various spice indicators are as follows; In the above formula, S k The standard deviation of the k-th indicator is represented. A k This represents the mean of the k-th indicator; 3) Calculation of weights (1) The importance score or coefficient of variation of different indicators is denoted as X 1. X 2…. X n Then the weight of the kth indicator a k The calculation is as shown in formula (1). a k Round to two decimal places: In the above formula, X k Let be the score or coefficient of variation of the k-th indicator; the weight calculated from the coefficient of variation is denoted as the objective weight 'a'. cv The weight calculated by expert scoring is denoted as subjective weight a. s ; (2) Final weight of each indicator a The weights are determined by both objective and subjective weights, and the calculation formula is as follows: 4) Determining the quality level and probability; The specific steps are as follows: Based on the actual situation, the spices are divided into several grades according to their quality, and each grade is numbered in order from best to worst, with the best grade being grade 1, and so on. The sample classification shall be determined according to the following criteria: (1) If all quality indicators involved in the grade classification reach or exceed a pre-defined critical point, then the sample meets the requirements for classification into that grade. (2) If at least one of the quality indicators involved in the grading fails to reach a predetermined critical point, the sample cannot be classified into that grade, but should be classified into the grade of the worst indicator. Here, the probability of a certain level occurring is defined. V The probability of a randomly selected spice sample falling into a specific category; Define the cumulative probability of level n P n For a given sample, the probability that all quality indicators involved in the grade classification reach or exceed the critical point specified by that indicator; Define the probability of each level occurring under random conditions. V The cumulative probability of level n P n The calculation is as shown in formula (4): In the formula, V i Let i be the probability of occurrence of level i. 5) Determination of the critical point for quality grading; The specific steps are as follows: Based on the probability corresponding to the quality level, determine the critical point for each indicator level by using weighting coefficients; 5-1) Determination of the probability of classifying the various evaluation indicators of spices Based on the overall evaluation index grading probability values ​​of existing spice samples, the grading probability requirements for each evaluation index are determined. For a quality index k, if a smaller observed value indicates better quality, then the cumulative probability of spice characteristic quality index k reaching grade n is determined. R kn Calculated using formula (5): For a quality index k, if a larger observed value indicates better quality, then the cumulative probability of the spice's characteristic quality index k reaching the grade n requirement is... R kn Calculated using formula (6): In the formula: R kn —The cumulative probability of the characteristic quality index k of spices meeting the requirements of grade n; P n —The cumulative probability of the nth level of spices; a k —The weight of the spice index k; 5-2) Determination of the critical points for each indicator classification Based on the cumulative probability of each indicator calculated above, and combined with the normal distribution curve of the indicator, we back-calculate its value under the cumulative probability, which serves as the critical value for grading. 6) Establish quality grading standards for spices based on grading thresholds.

2. The method for grading and evaluating the quality of spices as described in claim 1, characterized in that... Step 1) The relevant quality indicators of spices are: ； All the indicators are characterized by being "continuously distributed, quantifiable, and having a known positive or negative impact on quality"; among the "other" indicators, some indicators show results close to 100% in actual spice tests.

3. The method for grading and evaluating the quality of spices as described in claim 2, characterized in that... Step 2) The specific steps for selecting characteristic quality indicators are as follows: (1) Based on the part of the selected spice used, refer to the spice quality index table to determine the corresponding index; (2) Based on the function of the selected spices, select one indicator from each function in the spice quality-related index table for measurement; (3) Based on the characteristics of the selected spices, determine some of the "Other" indicators in the spice quality-related indicator table; (4) Based on the correlation analysis results of all the above-mentioned measurement index data, the final grading index is determined to ensure that there is no significant relationship between the retained quality indexes.

4. The method for grading and evaluating the quality of spices as described in claim 1, characterized in that... Step 3) The spice importance score evaluation table is as follows: 。 5. The method for grading and evaluating the quality of spices as described in claim 1, characterized in that... Step 6) Compile a grading standard based on the calculated critical points, and merge the parts whose critical points are too close to be measurable.

Images