Method for calculating indexes of meat beef and yak meat suitable for Chinese style grilling processing

Abstract

The invention discloses an index calculation method for meat beef and yak meat suitable for Chinese baking processing. The method comprises the following steps: respectively obtaining first physicochemical property index detection values of meat samples of different parts of a plurality of sample cattle before Chinese baking processing; respectively performing Chinese baking on the meat samples of different parts of the plurality of sample cattle according to a preset cooking condition, and obtaining a second physicochemical property index detection value of each meat sample after Chinese baking treatment; and based on the first physicochemical property index detection value and the second physicochemical property index detection value, calculating a comprehensive quality index value of each meat sample suitable for the Chinese baking cooking mode. By synchronously quantifying the raw material characteristics before grilling and the key texture indexes after grilling, a correlation model of cross-stage physicochemical characteristics is established, closed-loop prediction from the raw material characteristics to the terminal eating quality is realized, and the accuracy and universality of grilling suitability evaluation of different parts of different types of beef / yak meat are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of meat quality detection, and in particular to a method for calculating an index of beef and yak meat suitable for Chinese-style roasting. Background Art

[0002] Currently, the suitability evaluation of beef and yak beef for Chinese-style roasting relies primarily on traditional empirical judgment or single-dimensional analysis of physical and chemical indicators. In practice, processors often select meat parts based on market knowledge or historical culinary experience, for example prioritizing evenly distributed fat parts like eye ribs and top loin while avoiding tendons with dense connective tissue. This approach lacks a systematic correlation between the physical and chemical properties of the raw meat and textural changes after roasting, resulting in ambiguous selection criteria and low reproducibility.

[0003] Some methods use quantitative indicators to guide the processing process, but they mostly focus on the detection of isolated indicators. For example, they only detect the fat content or meat color value before grilling, believing that a high fat content (>5%) can enhance the flavor, or use a spectrophotometer to measure the L* value (brightness) to predict the color after grilling. Other methods only focus on a single texture characteristic after grilling, such as using a texture analyzer to measure the shear force value as a basis for tenderness evaluation, or detecting the thermal solubility of collagen to assess the degree of connective tissue transformation. Although this type of method introduces instrument detection, it ignores the dynamic coupling effect of multiple indicators during the grilling process, such as the synergistic effect of fat melting and collagen degradation on the final taste.

[0004] In addition, there are attempts in the existing technology to conduct comprehensive evaluations based on subjective weighting. For example, fixed weights are assigned to limited indicators such as fat content and tenderness through the Analytic Hierarchy Process (AHP), and a weighted comprehensive score is calculated. However, this method does not address the inherent conflict between indicators: certain high-fat cuts may increase their hardness during grilling due to insufficient collagen conversion, but static weight models cannot capture this nonlinear relationship. At the same time, traditional methods do not consider the redundancy of indicators. For example, cohesion and elasticity are often highly correlated, and direct superposition calculations will lead to distorted evaluation results.

[0005] What is particularly striking is that existing methods generally separate the association mechanism between pre-roasting characteristics and post-roasting quality. The high-temperature rapid cooking process of Chinese roasting can cause multiple changes such as protein denaturation, fat oxidation and Maillard reaction, but current technology has neither systematically quantified the transformation rules of key indicators before and after roasting (such as collagen thermal solubility and insoluble collagen residue), nor established an interaction model for cross-stage indicators. This has led to significant deviations in the evaluation of the suitability of high-collagen meats such as yak beef, making it difficult to accurately predict the roasting suitability of different parts. Summary of the Invention

[0006] The purpose of the embodiment of the present invention is to provide a method for calculating the index of beef and yak meat suitable for Chinese-style roasting processing. By synchronously quantifying the raw material characteristics before roasting (fat content, color) and the key texture indicators after roasting (collagen thermal solubility, insoluble collagen content, cohesion and elasticity), a correlation model of cross-stage physical and chemical properties is established, breaking through the limitations of traditional methods of isolated analysis of single-stage indicators. Based on the influence of fat distribution before roasting on heat transfer efficiency, the decisive role of collagen conversion degree after roasting on tenderness, and the contribution of viscoelastic parameters to chewing experience, a closed-loop prediction from raw material characteristics to terminal edible quality is achieved, which significantly improves the accuracy and universality of roasting suitability evaluation of different parts of different types of beef / yak meat.

[0007] To solve the above technical problems, a first aspect of an embodiment of the present invention provides a method for calculating an index of beef and yak meat suitable for Chinese-style roasting, comprising the following steps:

[0008] Obtaining first physical and chemical property index test values ​​of meat samples from different parts of several sample cattle before Chinese-style grilling, wherein the first physical and chemical property indexes include fat content and raw meat color value;

[0009] After Chinese-style grilling of several sample cattle meat samples from different parts according to preset cooking conditions, a second physical and chemical property index test value of each meat sample after Chinese-style grilling is obtained, wherein the second physical and chemical property index includes: thermal solubility of grilled meat collagen, content of insoluble collagen in grilled meat, cohesion and elasticity;

[0010] Based on the detection values ​​of the first physicochemical property index and the second physicochemical property index, a comprehensive quality index value of each meat sample suitable for Chinese-style grilling cooking method is calculated.

[0011] Furthermore, the calculation of the comprehensive quality index value of each meat sample suitable for Chinese-style grilling cooking method based on the first physical and chemical property index detection value and the second physical and chemical property index detection value includes:

[0012] Normalizing the detection values ​​of each of the first physicochemical property indicators and each of the second physicochemical property indicators, respectively, and obtaining purification index sets for each of the first physicochemical property indicators and each of the second physicochemical property indicators through three-level joint screening;

[0013] Performing a dual-path fusion calculation on the purified indicator set to obtain a final weight set of the purified indicator set, wherein the dual paths include: a principal component analysis path for extracting initial weights of factor loads of principal components with high contribution rates, and a variation correlation path for calculating indicator conflict coefficients;

[0014] Performing a weighted sum calculation on the normalized data of the purification index set and the final weight set to obtain a comprehensive quality index value of each meat sample;

[0015] The calculation formula of the comprehensive quality index value IQI is:

[0016]

[0017] Among them, n is the number of indicators contained in the purification indicator set, ω i is the final weight of the i-th indicator, N i is the normalized value of the i-th indicator detection value.

[0018] Furthermore, the purification index set of each of the first physical and chemical property indexes and each of the second physical and chemical property indexes obtained by the three-level joint screening includes:

[0019] Calculating a Pearson correlation coefficient matrix of each of the first physicochemical property indices and each of the second physicochemical property indices based on the normalized detection values ​​of each of the first physicochemical property indices and each of the second physicochemical property indices, and eliminating redundant indices with correlations higher than a preset threshold in the Pearson correlation coefficient matrix to obtain a preliminary screening indicator set;

[0020] Based on the preliminary screening index set, calculating the coefficient of variation of each of the first physicochemical property indexes and each of the second physicochemical property indexes, eliminating low-sensitivity indexes whose coefficient of variation is lower than a preset coefficient of variation threshold, and obtaining an intermediate screening index set;

[0021] Based on the intermediate screening indicator set, the Spearman rank correlation coefficient of every two indicators is calculated and the Spearman rank correlation coefficient matrix of the intermediate screening indicator set is constructed, and independent indicators whose correlation with other indicators is lower than a preset correlation threshold are retained to obtain a purified indicator set.

[0022] Furthermore, performing dual-path fusion calculation on the purification index set to obtain a final weight set of the purification index set includes:

[0023] Based on the purification index set, principal component analysis processing is performed, including performing eigenvalue decomposition on the purification index set matrix, extracting principal components whose cumulative variance contribution rate reaches a preset standard, determining the factor loading of each of the first physical and chemical property indicators and each of the second physical and chemical property indicators through eigenvectors, and obtaining an initial weight set of principal components;

[0024] Based on the purification index set, calculating the coefficient of variation of each of the first physical and chemical property indicators and each of the second physical and chemical property indicators and the rank correlation coefficient with other indicators, and calculating the indicator conflict coefficient using the CRITIC weight method to obtain a conflict weight set;

[0025] Based on the initial weight set of the principal components and the conflicting weight set, weighted fusion and normalization processing are performed to generate a final weight set of the purification index set.

[0026] Furthermore, the first physical and chemical property index test values ​​of the meat samples of different parts of several sample cattle before Chinese-style roasting are obtained respectively, including:

[0027] Obtaining a fat content detection value of the meat sample before Chinese-style grilling by Soxhlet extraction and / or near-infrared spectroscopy;

[0028] The raw meat color detection value of the meat sample before Chinese-style roasting is obtained by spectrophotometry.

[0029] Furthermore, the obtaining of the second physical and chemical property index detection value of each meat sample after Chinese-style roasting includes:

[0030] Obtaining the thermal solubility test value of collagen in the meat sample after Chinese-style roasting by thermal extraction-colorimetry;

[0031] Obtaining the insoluble collagen content of the meat sample after Chinese-style roasting by acid hydrolysis and centrifugation;

[0032] Obtaining a cohesiveness test value of the meat sample after Chinese-style roasting by texture profile analysis;

[0033] The elasticity detection value of the meat sample after Chinese-style roasting is obtained by texture profile analysis.

[0034] Furthermore, the meat sample includes eye meat, top meat, shoulder meat, shank, rump meat, tenderloin and / or breast meat of the sample beef.

[0035] Furthermore, the sample cattle include beef cattle and yaks.

[0036] Accordingly, the second aspect of an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the above-mentioned method for calculating the index of beef and yak meat suitable for Chinese-style roasting processing.

[0037] Accordingly, a third aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implements the above-mentioned method for calculating the index of beef and yak meat suitable for Chinese-style roasting.

[0038] The above technical solutions of the embodiments of the present invention have the following beneficial technical effects:

[0039] 1. By integrating pre-grilled raw material characteristics (fat content, color) with post-grilled texture indicators (collagen thermal solubility, insoluble collagen content, cohesiveness, and elasticity), a comprehensive quality index across processing stages is constructed. This method breaks through the limitations of traditional methods that analyze single-stage indicators in isolation, quantifies the role of fat melting in promoting heat transfer efficiency and the decisive influence of collagen thermal conversion on tenderness, and achieves closed-loop modeling of the path of grilled quality formation. This model significantly improves the adaptability prediction accuracy of highly variable parts (such as yak shank and tendon meat);

[0040] 2. A purified index set is constructed based on three-level joint screening (Pearson correlation to eliminate redundancy, coefficient of variation to retain sensitivity, and Spearman rank correlation to extract independence). Combined with dual-path fusion weight calculation (principal component analysis to extract high-contribution factors and the CRITIC method to quantify index conflicts), the final weight set is dynamically generated. This method effectively resolves the inherent conflicts in indicators used in traditional static weighting (such as the antagonistic effect of high fat content and low collagen solubility), enabling the comprehensive index to more accurately reflect complex physical and chemical interaction mechanisms.

[0041] 3. Data comparability is achieved through a standardized chain of testing methods. First, before roasting, Soxhlet extraction / near-infrared spectroscopy is used to measure fat and a spectrophotometer is used to determine color. Second, after roasting, collagen solubility is measured by thermal extraction-colorimetry, insoluble collagen is measured by acid hydrolysis and centrifugation, and cohesion / elasticity is determined by texture profile analysis. Third, systematic sampling of multiple parts (7 categories, including eye meat and shoulder meat) and normalization are combined. This standardized framework ensures traceability of data from raw material characteristics to final texture, eliminates interference from human experience (such as subjective bias in traditional roasting trials), and establishes a universal evaluation benchmark for different varieties / parts of beef / yak beef. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of a method for calculating an index of beef and yak meat suitable for Chinese-style roasting provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present invention.

[0044] Please refer to Figure 1 A first aspect of the embodiments of the present invention provides a method for calculating an index of beef and yak meat suitable for Chinese-style roasting, comprising the following steps:

[0045] Step S100 , obtaining first physical and chemical property index test values ​​of several types of meat samples from different parts of cattle before Chinese-style roasting, wherein the first physical and chemical property indexes include fat content and raw meat color value.

[0046] Step S100 is mainly aimed at testing the basic physical and chemical properties of raw meat; fat content is the core indicator, which directly affects the heat conduction efficiency and flavor substance production during the roasting process - the melting behavior of fat cells at high temperatures can promote the deep penetration of heat into muscle fibers, and at the same time participate in the Maillard reaction to generate characteristic roasted aroma. The color value of raw meat (usually represented by CIE L*a*b* color space) reflects the state of myoglobin and surface light scattering properties. Its initial value not only determines the color potential after roasting, but is also indirectly related to the pH value and water holding capacity of the muscle. By measuring fat through Soxhlet extraction / near-infrared spectroscopy and capturing color with a spectrophotometer, a quantifiable raw material benchmark database is established to provide an accurate data basis for subsequent cross-stage correlation analysis.

[0047] In step S200, after Chinese-style roasting of meat samples from different parts of several sample cattle according to preset cooking conditions, a second physical and chemical property index detection value of each meat sample after Chinese-style roasting is obtained, and the second physical and chemical property index includes: thermal solubility of roasted meat collagen, content of insoluble collagen in roasted meat, cohesion and elasticity.

[0048] The Chinese-style grilling process induces significant microstructural reorganization of meat. Step S200 captures key textural changes using standardized testing methods. Collagen thermal solubility directly reflects the efficiency of connective tissue conversion. Under high temperature, the collagen triple helix unwinds and forms gelatin. The solubility rate is significantly positively correlated with tenderness. The extent of this conversion depends on the efficiency and duration of heat penetration. For example, fully dissolved collagen enhances the smoothness of the meat. Meanwhile, the insoluble collagen content serves as an inverse indicator, indicating the amount of recalcitrant collagen residue that has not fully gelled. This tightly cross-linked collagen network strengthens the mechanical connection between muscle bundles. Experimental data show that a 1% increase in insoluble collagen content increases shear force by an average of 8-12 N. This metric provides a direct basis for evaluating the effects of the grilling process on connective tissue. Cohesion and elasticity jointly describe the macroscopic mechanical properties of denatured proteins. Cohesion characterizes the cohesive strength between muscle fiber fragments, reflecting the meat's ability to resist delamination and fragmentation during chewing. Elasticity quantifies the degree of deformation recovery after compression and is closely related to the degree of thermal denaturation of myofibrillar proteins and water migration. These two parameters can be accurately obtained through texture profile analysis (TPA), avoiding the limitations of vague descriptions such as "tough but not woody" in traditional sensory evaluation.

[0049] Step S300: Calculate the comprehensive quality index value of each meat sample suitable for Chinese-style grilling cooking method based on the first physical and chemical property index detection value and the second physical and chemical property index detection value.

[0050] Step S300 constructs a mathematical correlation model between pre-roasting raw material properties and post-roasting texture response. By integrating the effect of fat content on heat conduction (fat melting increases thermal conductivity by approximately 30%) and the contribution of collagen thermal solubility to tenderness, a quantitative transmission chain from "raw material properties to processing response" is formed. This cross-stage integration overcomes the limitations of traditional methods that analyze single indicators in isolation.

[0051] The resulting comprehensive index offers multi-dimensional characterization: it incorporates mechanical properties (cohesion and elasticity) to predict chewing experience, biochemical indicators (collagen conversion rate) to reflect tenderness development efficiency, and correlates the impact of the Maillard reaction on visual quality through pre-grill color. This model establishes a quantifiable benchmark for grilling suitability evaluation for different types of beef and anatomical parts, such as collagen-rich yak shank.

[0052] Furthermore, in step S300, based on the first physicochemical property index detection value and the second physicochemical property index detection value, the comprehensive quality index value of each meat sample suitable for Chinese-style grilling cooking method is calculated, including:

[0053] In step S310 , normalizing the detection values ​​of each first physicochemical characteristic index and each second physicochemical characteristic index is performed respectively, and obtaining purification index sets of each first physicochemical characteristic index and each second physicochemical characteristic index through three-level joint screening.

[0054] Step S310 is used to solve the problem of information redundancy and low discrimination between multiple source indicators. First, the original test values ​​such as fat content before grilling, raw meat color value, and collagen thermal solubility, insoluble collagen content, cohesion, and elasticity after grilling are normalized to eliminate the interference of different dimensions (such as the magnitude difference between fat content percentage and color difference value L*) on subsequent analysis. Then, a three-level joint screening mechanism is implemented: the Pearson correlation coefficient matrix is ​​used to identify highly correlated indicator groups (such as cohesion and elasticity often show a strong positive correlation), and redundant indicators are eliminated according to a preset threshold (such as |r|>0.85), retaining representative parameters that are more explanatory of grilling quality; the data volatility of the remaining indicators is analyzed based on the coefficient of variation, and low-sensitivity indicators with a coefficient of variation below the threshold (such as <0.15) are eliminated to ensure that the screened indicators can effectively distinguish the characteristic differences of meat from different parts; finally, the Spearman rank correlation coefficient matrix is ​​constructed to retain indicators whose correlation with other indicators is lower than the independence threshold (such as |ρ|<0.3) to avoid collinearity interference. The purification indicator set generated by the three-level screening has low redundancy, high discrimination and statistical independence, laying the data foundation for weight calculation.

[0055] Step S320, performing a dual-path fusion calculation on the purified indicator set to obtain the final weight set of the purified indicator set, wherein the dual paths include: a principal component analysis path for extracting the initial weights of the factor loads of the principal components with high contribution rates, and a variation correlation path for calculating the indicator conflict coefficient.

[0056] Step S320 implements scientific weighting by integrating statistical contribution and indicator conflict. The principal component analysis pathway focuses on the explanatory power of indicators for overall variation: Eigenvalue decomposition is performed on the purified indicator set matrix to extract principal components whose cumulative variance contribution meets a preset standard (e.g., ≥85%). The factor loadings of each indicator on the core principal component are calculated using eigenvectors (e.g., collagen thermal solubility can have a loading of 0.91 on the first principal component). After normalization, an initial weight set is formed. This pathway quantifies the dominant role of key indicators in overall quality, such as how fat content significantly correlates with multiple principal components by affecting heat conduction efficiency. The variation correlation pathway, based on the CRITIC weighting method, simultaneously calculates the coefficient of variation (reflecting data dispersion) and the conflict coefficient (comprised of 1 minus the absolute value of the Spearman rank correlation coefficient between indicators). The conflict coefficient quantifies the antagonistic relationship between indicators (e.g., the color value of raw meat and the elasticity of grilled meat may be negatively correlated due to the myoglobin denaturation mechanism), thus avoiding evaluation bias caused by ignoring indicator conflicts in traditional weighting. The dual-path weights are weighted and fused (e.g., the principal component weight accounts for 60% and the conflict weight accounts for 40%) and normalized to form the final weight set. This method retains high-contribution indicators while explicitly resolving the dynamic constraints between them.

[0057] Step S330 , performing weighted sum calculation on the normalized data of the purification index set and the final weight set to obtain a comprehensive quality index value of each meat sample.

[0058] The calculation formula of the comprehensive quality index value IQI is:

[0059]

[0060] Among them, n is the number of indicators contained in the purification indicator set, ω i is the final weight of the i-th indicator, N i is the normalized value of the i-th indicator detection value.

[0061] Step S330 is the core calculation link of the entire Chinese-style barbecue meat quality evaluation process. Its goal is to aggregate the indicator information that has been strictly screened and weighted in the early stage into an intuitive, quantitative comprehensive evaluation value, namely the comprehensive quality index (IQI). This step directly acts on the purified indicator set obtained in step S310 (including the key physical and chemical indicators with low redundancy, high discrimination and statistical independence retained after three-level screening) and the final weight set calculated in step S320 (the weight set integrates the contribution of the indicators to the overall variation and their conflicts); the applicable scenario clearly points to the need for rapid, objective and quantitative evaluation of the comprehensive performance of the expected quality of different parts or batches of fresh meat raw materials after Chinese-style barbecue cooking, providing core data support for meat grading, raw material screening, process optimization or consumer quality prediction.

[0062] For each indicator in the purification index set, take its corresponding normalized detection value (N i , the value has eliminated the original dimension difference and is in a comparable numerical range, usually mapped to the interval [0,1]), and multiplied by the weight (ω) corresponding to the indicator in the final weight set determined in step S320 i Then, the weighted result (ω) of all indicators (the total number of indicators is recorded as n) is i ×N i ) are accumulated and summed to obtain the comprehensive quality index (IQI) of the meat sample.

[0063] First, it is an integrated application of the results of previous complex data processing: normalized value (N i ) ensures that indicators of different properties and dimensions (such as fat percentage, color difference value, texture parameters) can be integrated on a fair scale; the purified indicator set (n indicators) ensures the validity and representativeness of the input information and eliminates noise and redundancy; the final weight set (ω i) scientifically quantifies the relative importance of each indicator to the overall quality of grilling. The weight not only takes into account the indicator's own ability to explain data variation (through the principal component analysis path), but also takes into account the possible antagonistic or conflicting relationships between indicators (through the variation-related path), thereby avoiding the bias caused by the traditional weighting method that may ignore the dynamic interaction between indicators. Secondly, the calculated IQI value is designed to be in the range of [0,1]. The higher the value, the more suitable the meat sample is for Chinese grilling cooking methods at the level of physical and chemical properties, and it is expected to obtain better overall quality (such as ideal tenderness, juiciness, flavor precursors, color stability, etc.). This single scalar output format greatly simplifies the interpretation of complex multidimensional data, and can quickly and intuitively compare the grilling suitability of different samples, providing a direct and reliable quantitative basis for subsequent precise applications (such as high-end grilled meat screening, product formula adjustment, and cooking parameter guidance). Ultimately, IQI, as a comprehensive indicator that integrates multiple dimensions of key physical and chemical properties, effectively solves the one-sided problem that may exist in evaluating roasting quality by relying on only a single or a few indicators, and realizes a holistic and scientific assessment of the suitability of meat for roasting.

[0064] For example, if a high fat content sample is accompanied by high collagen thermal solubility (ω 脂肪 With ω 热溶解性 weights are large), the index value increases significantly; on the contrary, if the high-fat sample has insoluble collagen residue (ω 不溶性胶原 The model automatically captures the synergistic and antagonistic effects between indicators through weight assignment (for example, the promotion of collagen conversion by fat is implicit in the weight association). The final output directly reflects the degree of suitability of different meat samples for Chinese grilling.

[0065] Furthermore, the purification index set of each first physical and chemical property index and each second physical and chemical property index obtained through the three-level joint screening in step S310 includes:

[0066] Step S311, based on the normalized detection values ​​of each first physicochemical property index and each second physicochemical property index, calculate the Pearson correlation coefficient matrix of each first physicochemical property index and each second physicochemical property index, eliminate redundant indicators with correlation higher than a preset threshold in the Pearson correlation coefficient matrix, and obtain a preliminary screening indicator set.

[0067] The calculation formula of Pearson correlation coefficient is:

[0068]

[0069] Among them, r xy is the Pearson correlation coefficient between indicator x and indicator y, x s is the normalized detection value of the s-th meat quality sample on index x, is the mean value of index x in all meat samples, y s is the normalized detection value of the s-th meat sample on index y, is the mean of index y in all meat samples, and m is the number of all meat samples.

[0070] Step S311 is the first stage of the three-level joint screening mechanism, which is used to solve the information redundancy problem between multi-source physical and chemical indicators; it acts on all the first and second physical and chemical characteristic indicator test values ​​after normalization (such as fat content and meat color L* value before grilling, collagen thermal solubility and cohesion after grilling, etc.). By quantifying the linear correlation strength between indicators, highly correlated repeated or redundant indicators are identified and eliminated to avoid the same information being counted multiple times, resulting in distortion of subsequent analysis. In terms of implementation method, the Pearson correlation coefficient (r) between all indicators is first constructed. xy ) matrix. The Pearson correlation coefficient formula accurately calculates the linear correlation degree of any two indicators x and y, where x s and y s is the normalized value of the corresponding index of the i-th meat quality sample, and is the mean of each indicator in all m samples. Based on the preset strong correlation threshold (such as |r|>0.85), highly correlated indicator groups are identified (for example, cohesion and elasticity often show a strong positive correlation). Subsequently, according to preset rules (such as retaining indicators that are more closely related to the quality of the roasted core or are easier to measure), redundant indicators are eliminated from each group, and only one representative indicator is retained. The above processing method effectively compresses the indicator dimensions, eliminates the information overlap caused by high collinearity, and generates a preliminary screening indicator set with higher information density and stronger representativeness, laying a low-redundancy data foundation for subsequent screening.

[0071] Step S312: Based on the preliminary screening index set, the coefficient of variation of each first physicochemical property index and each second physicochemical property index is calculated, and low-sensitivity indexes with a coefficient of variation lower than a preset coefficient of variation threshold are eliminated to obtain an intermediate screening index set.

[0072] Coefficient of variation (CV) i The calculation formula is:

[0073]

[0074] Among them, σ i is the standard deviation of the detection value of the i-th indicator, μ i is the mean of the detection values ​​of the i-th indicator.

[0075] Step S312 continues the preliminary screening index set to solve the problem of insufficient discrimination of the remaining indicators, that is, to ensure that the retained indicators can effectively distinguish the differences in the characteristics of meat from different parts or batches. Focus on quantifying the relative dispersion of each indicator between samples and eliminate those "low-sensitivity" indicators that have too small fluctuations between different samples and lack the ability to distinguish. Calculate the coefficient of variation (CV) of each i-th indicator in the preliminary screening index set. i ). The coefficient of variation formula is a dimensionless statistic, where σ i is the standard deviation of index i, reflecting the absolute dispersion of the data; μ i is the mean of indicator i. CV i By normalizing the standard deviation to a ratio of the mean, the influence of the magnitude or unit of the indicator itself on the dispersion comparison is eliminated, so that the volatility of indicators of different properties (such as percentage content and instrument measurement value) can be directly compared. i <0.15, and those indicators with too low coefficient of variation are eliminated. For example, if the thermal solubility of collagen protein in a certain part of meat is very close in all samples (such as high mean and very small standard deviation, resulting in CV i If the sensitivity of the selected meat samples is very low, their ability to distinguish between different meat qualities for roasting is weak and they should be eliminated. This step ensures that the final evaluation index has sufficient data volatility and sensitivity to capture the substantial differences between different meat samples, thus generating an intermediate screening index set with greater discriminatory power.

[0076] Step S313, based on the intermediate screening indicator set, calculate the Spearman rank correlation coefficient between every two indicators and construct the Spearman rank correlation coefficient matrix of the intermediate screening indicator set, retain independent indicators whose correlation with other indicators is lower than the preset correlation threshold, and obtain the purified indicator set.

[0077] The calculation formula of Spearman rank correlation coefficient is:

[0078]

[0079] Among them, ρ xy is the Spearman rank correlation coefficient between indicator x and indicator y, d s is the rank difference of the sth sample of the two indicators.

[0080] Step S313 is the final stage of the three-level screening, which acts on the intermediate screening indicator set to solve the potential collinearity problem between indicators and ensure the statistical independence of the final indicator set, avoiding the interference of residual nonlinear or monotonic correlations in subsequent weight calculations (such as principal component analysis); focusing on the possible nonlinear but monotonic correlations between indicators. By constructing the Spearman rank correlation coefficient (ρ) between all the indicators in the intermediate screening indicator set, the results are summarized. xy ) matrix. The Spearman correlation coefficient formula calculates the Pearson correlation coefficient between the ranks of indicators x and y (i.e., the ranking position of each indicator value in all samples, such as the maximum rank is m and the minimum rank is 1). s Is the rank difference of the sth sample on the two indicators. Spearman coefficient ρ xy It measures the consistency of the changing trends of the two indicators (monotonic relationship), and is sensitive to nonlinear but monotonic relationships (such as an increase in the color value L* of raw meat may cause the elasticity of grilled meat to show a monotonically decreasing trend). After the calculation is completed, based on the preset independence threshold (such as requiring that the |ρ| of a certain indicator and all other indicators is less than 0.3), those indicators with generally low correlation with other indicators are retained. The above steps further refine the indicator set and minimize the statistical dependence (collinearity) between indicators. The final purified indicator set not only has low redundancy (from S311) and high discrimination (from S312), but also has good statistical independence. This independence provides a reliable data basis for the scientific calculation of indicator weights (especially the principal component analysis path) in the subsequent step S320, ensuring that the weights can truly reflect the unique contribution of each indicator to the overall quality, rather than being interfered with by other related indicators.

[0081] Furthermore, in step S320, the dual-path fusion calculation is performed on the purification index set to obtain the final weight set of the purification index set, including:

[0082] Step S321: Perform principal component analysis based on the purification index set, including performing eigenvalue decomposition on the purification index set matrix, extracting principal components whose cumulative variance contribution rate reaches a preset standard, determining the factor loading of each first physicochemical property index and each second physicochemical property index through the eigenvector, and obtaining the initial weight set of the principal components.

[0083] The calculation formula for the initial weight of the principal component is:

[0084]

[0085] Among them, v ir is the component of the eigenvector of the rth principal component on the i-th index, λ r is the eigenvalue of the rth principal component, and k is the number of principal components retained.

[0086] Step S321 is the first path of the dual-path weighting mechanism, which is used to quantify the contribution of each indicator in the purification index set to the overall data variation. It acts on the purification index set after three-level screening (including p low-redundancy, high-discrimination and independent physical and chemical indicators); it is used to identify the core indicators that have the most significant impact on the overall quality of the grilling, avoiding subjective weighting bias. First, the m×p-dimensional data matrix of the purification index set is subjected to eigenvalue decomposition, and the eigenvalues ​​of the covariance matrix (λ r ) and the corresponding eigenvector (component v ir ). Eigenvalue λ r The size of directly reflects the ability of the principal component to explain the variation of the original data. It is sorted from large to small by eigenvalue, and the top k principal components whose cumulative variance contribution rate reaches the preset standard (e.g. ≥85%) are extracted to ensure that they can represent the vast majority of information. Subsequently, the initial weight of each indicator i is calculated using the formula: where |v ir ∣ is the absolute value of the factor loading of the i-th indicator on the r-th principal component (reflecting the strength of the association between the indicator and the principal component), λ r is the variance contribution weight of the principal component. If an indicator has a high load on multiple high-variance principal components (such as collagen thermal solubility with a load of 0.91 on the first principal component), its contribution to the overall variation is greater. Finally, the initial weights are normalized to generate an initial set of principal component weights. From a data-driven perspective, the above path objectively identifies key indicators that play a dominant role in the variation in grilling quality (such as fat content, which is significantly associated with multiple principal components by affecting heat conduction efficiency), providing a solid statistical basis for the final weights.

[0087] Step S322: Based on the purification index set, the coefficient of variation of each first physicochemical property index and each second physicochemical property index and the rank correlation coefficient with other indicators are calculated, and the index conflict coefficient is calculated using the CRITIC weight method to obtain a conflict weight set.

[0088] The calculation formula of CRITIC conflict weight is:

[0089]

[0090] Among them, C i is the information amount of the i-th indicator, j is the other indicators in the purification indicator set except i, n is the total number of indicators in the purification indicator set, r ij is the Pearson correlation coefficient between the i-th indicator and the j-th indicator in the purification indicator set.

[0091] Step S322 is the second path of the dual-path weighting, which is used to quantify the conflict (antagonistic relationship) between indicators and integrate data volatility to solve the problem of ignoring the dynamic interaction of indicators in traditional weighting. The above path acts on the same purification indicator set. In view of the mutual restriction of physical and chemical properties that may exist in the roasting process (such as the negative correlation mechanism between meat color value and elasticity), the CRITIC (Criteria Importance Through Intercriteria Correlation) weighting method is adopted. First, the coefficient of variation (CV) of each indicator i is calculated. i ), whose formula reflects the relative dispersion (discrimination) of this indicator among different meat samples. Secondly, calculate the indicator conflict coefficient: by constructing the Pearson correlation coefficient matrix between indicators to quantify the information conflict between indicator i and all other indicators. ij The closer || is to 1, the higher the information overlap and the lower the conflict between indicators. ij On the contrary, the larger the value, the more independent the information provided by indicator i is (for example, when an indicator is weakly correlated with other indicators, its conflict coefficient is higher). Finally, the comprehensive information content C of indicator i is i It is determined by the product of the coefficient of variation and the conflict coefficient. This value also takes into account the indicator's own discrimination (CV i ) and the independent contribution of indicators After normalization, conflicting weights are obtained, which explicitly deal with the antagonistic or complementary relationship between indicators (such as avoiding high weights while giving strong negative correlation indicators), ensuring that the final weight set can reflect the dual value of the indicator's information uniqueness and data discrimination.

[0092] In step S311, the symbol r xy Represents the Pearson correlation coefficient, which is used to quantify the linear correlation strength between any two indicators x and y in the original unfiltered indicator set and to identify highly correlated indicator groups (such as |r xy |>0.85), and solve the information overlap problem between multiple source indicators by eliminating redundant indicators (such as choosing between cohesion and elasticity). This stage acts on all initial indicators to build a low-redundancy preliminary screening indicator set for subsequent analysis. In step S322, the symbol r ij It also represents the Pearson correlation coefficient, but it only works on the purified indicator set after the three-level screening. Where i represents the current target indicator, and j traverses all indicators except i. Its core purpose is to calculate the correlation coefficient r of indicator i with other indicators. ij , and then quantify the independent information contribution of the indicator through the conflict coefficient (such as |r ij ∣≈0.2, the conflict is high). This result is used in the CRITIC weight calculation to ensure that the weight distribution takes into account both the discrimination and independence of indicators.

[0093] Step S323 : performing weighted fusion and normalization processing based on the initial weight set of the principal components and the conflicting weight set to generate a final weight set of the purification index set.

[0094] The final weight is calculated as follows:

[0095]

[0096] Among them, θ1 and θ2 are fusion coefficients.

[0097] Step S323 is the integration phase of dual-path weighting. Its core goal is to fuse the statistical contribution of the principal component path with the information independence of the conflicting paths to form scientific weights that are both explanatory and robust. The inputs to step S323 are the initial principal component weights generated in S321 and the conflicting weights generated in S322. The weights must simultaneously reflect the objective impact of the indicator on quality variation and its irreplaceable role in the evaluation system. Weighted linear fusion is employed: for each indicator i, a fusion weight is calculated. Here, θ1 and θ2 are preset fusion coefficients (by default, θ1 = θ2 = 0.5), satisfying θ1 + θ2 = 1 to ensure consistent weighting. The coefficient ratio can be adjusted based on actual needs (e.g., increasing θ1 to focus on dominant factors). Principal component weights emphasize the global influence of an indicator in the multidimensional variation space (e.g., fat content gains a high weight by correlating with multiple principal components), while conflicting weights enhance the incremental value of an indicator's local information (e.g., a low-correlation indicator receives an increased weight due to its unique information). Finally, the fusion weights are normalized to generate the final set of weights used to calculate the integrated quality index (IQI). By balancing the ability of indicators to explain overall variation with the dynamic conflict between indicators, the method avoids the limitations of single weighting methods (such as PCA, which ignores indicator antagonism, and CRITIC, which ignores absolute contributions). This ensures that the final weights capture core driving factors while also ensuring the comprehensiveness and robustness of the evaluation system, providing a core guarantee for the scientific nature and reliability of IQI calculation.

[0098] In a specific implementation of the embodiment of the present invention, step S100 of respectively obtaining the first physical and chemical property index test values ​​of meat samples from different parts of several sample cattle before Chinese-style grilling includes:

[0099] Step S110 , obtaining a fat content detection value of the meat sample before Chinese-style grilling by Soxhlet extraction and / or near-infrared spectroscopy.

[0100] Accurate quantification of fat content utilizes complementary technical approaches. Soxhlet extraction, the benchmark method, separates free and bound fat from muscle tissue through ether extraction cycles. After solvent evaporation, the residual lipid mass is weighed and the absolute fat content (in g / 100g) is calculated. While this method is time-consuming (approximately 8 hours per sample), it offers the same accuracy as destructive testing. Near-infrared spectroscopy provides a non-contact, rapid prediction model. This utilizes the absorption properties of C-H bonds in fat molecules at characteristic wavelengths such as 1200nm and 1730nm, combined with a pre-established calibration curve (based on a Soxhlet calibration sample set), enabling non-destructive testing within 60 seconds. The choice between the two methods depends on a balance between accuracy and efficiency: Soxhlet is used for laboratory calibration and verification, while near-infrared spectroscopy is suitable for batch screening on the production line. The key to fat data lies in its significantly higher thermal conductivity (0.17-0.21 W / m·K) than muscle fibers (0.45 W / m·K), which directly impacts the heat penetration efficiency during the grilling process.

[0101] Step S120 , obtaining the color detection value of the raw meat of the meat sample before Chinese-style grilling by using a spectrophotometer method.

[0102] To obtain the initial color of different yak beef cuts before Chinese-style grilling and achieve high-precision measurement of meat color, a key sensory and quality indicator, this embodiment explicitly stipulates the use of spectrophotometry as a standardized testing method. This relies on specialized spectrophotometer equipment. Its operating principle is to illuminate the fresh surface of a smooth-cut meat sample with a standard light source (typically a D65 light source that simulates daylight) and accurately measure the spectral composition of the light reflected from the sample surface. The instrument's built-in optical sensor and spectral analysis system convert the captured reflectance spectrum information into an internationally recognized, quantitative color space. The most commonly used is the CIE Lab (or L*a*b*) color space system. In this three-dimensional color space, the L* value represents the sample's lightness (brightness, ranging from 0 black to 100 white), the a* value represents the redness and greenness (positive values ​​indicate redness, negative values ​​indicate greenness), and the b* value represents the yellowness and blueness (positive values ​​indicate yellowness, negative values ​​indicate blueness). Therefore, each meat sample can obtain a set of accurate L*, a*, and b* values ​​before grilling, which fully characterizes its initial color characteristics, especially the key redness (a* value) that reflects the state and distribution of myoglobin in the muscle. The effect of using spectrophotometer is significant and has multiple advantages. First, it completely eliminates the subjective judgment differences and human errors existing in traditional visual colorimetry (such as colorimetric plate controls), ensuring the objectivity and repeatability of color evaluation. Secondly, this technology provides extremely high accuracy and sensitivity, and can capture subtle changes in meat color that are difficult to distinguish with the naked eye. This is crucial for studying the inherent differences in meat quality in different parts or evaluating slight processing effects. Third, the measurement process is fast and non-destructive (or only requires a very small sample surface), allowing continuous tracking of the same batch of samples or obtaining multi-point data on precious samples, greatly improving experimental efficiency and data richness. Ultimately, the L*, a*, and b* data obtained by the spectrophotometer provide a solid, reliable, and standardized basic data support for analyzing the influence mechanism of the roasting process (temperature, time) on meat color changes (such as browning and coking) in subsequent studies of this invention, as well as establishing the correlation between the initial color of the raw meat and the appearance quality of the finished product at the end of roasting.

[0103] Furthermore, obtaining the second physical and chemical property index detection value of each meat sample after Chinese-style grilling in step S200 includes:

[0104] Step S210, obtaining the thermal solubility test value of collagen of the meat sample after Chinese-style roasting by thermal extraction-colorimetry.

[0105] The thermal denaturation properties of collagen are used to quantify the degree of its conversion during the grilling process. The high temperature of Chinese grilling (usually >150°C) causes the collagen in the muscle connective tissue to undergo thermal depolymerization, partially converting it into soluble gelatin. The thermal extraction method simulates the thermal environment of cooking by continuously heating the grilled meat sample at a controllable temperature (such as a 70°C water bath) to fully dissolve the heat-sensitive collagen in the extract (usually a neutral buffer). The soluble collagen (gelatin) content in the extract is indirectly determined by colorimetry - based on the specific binding reaction between gelatin and a specific dye (such as a hydroxyproline color developer), the absorbance is measured at a specific wavelength (such as 560nm), and the absorbance is compared with a standard curve and converted to the soluble collagen content (unit: mg / g tissue). This test value directly reflects the proportion of collagen in the meat that has been softened by heat after grilling. Its principle is that the thermal solubility of collagen is significantly positively correlated with the tenderness of the meat. Through this method, the degree of damage to the connective tissue structure caused by the roasting process can be objectively quantified, providing key biochemical basis for analyzing the tenderization mechanism.

[0106] Step S220 , obtaining a detection value of the insoluble collagen content of the meat sample after Chinese-style roasting by acid hydrolysis and centrifugation.

[0107] Determination of the content of collagen that remains in a stable fiber form after grilling, based on the tolerance of insoluble collagen to strong acid treatment: After the meat sample is grilled at high temperature, the remaining undissolved collagen fibers need to be hydrolyzed by strong acid (such as 6M HCl) to destroy their cross-linked structure and release the characteristic amino acid hydroxyproline. After the acid hydrolysis product is centrifuged, the supernatant is reacted with a color developer, and the hydroxyproline content is also determined by colorimetry (such as 550nm wavelength), and then converted to the total insoluble collagen content (unit: mg / g tissue) according to a fixed coefficient (usually hydroxyproline accounts for 13.5% of the total collagen). This method focuses on the collagen component that still maintains rigidity after high-temperature treatment, and its content is closely related to the residual toughness and chewing resistance of the meat. Combined with the solubility data of S210, the overall transformation efficiency of the collagen network caused by grilling can be fully evaluated, revealing the differences in structural stability of meat from different parts during thermal processing.

[0108] Step S230: obtaining a cohesiveness test value of the meat sample after Chinese-style roasting by texture profile analysis.

[0109] Cohesion characterizes the ability of the internal structure of meat to resist destruction and maintain integrity, and is one of the core parameters of texture profile analysis (TPA). It is achieved through a physical property tester: the roasted meat sample cooled to room temperature is compressed twice by a cylindrical probe (the compression ratio is usually set to 50%), and the instrument records the stress-strain curve. The cohesion value is defined as the ratio of the positive peak area of ​​the second compression to the first compression (dimensionless). Its principle reflects the ability of the muscle fibers and connective tissue network to restore the original binding force of the meat sample after the first compression. High cohesion indicates that the meat can maintain structural coherence during chewing and is not easy to break; low values ​​indicate loose tissue. This parameter is directly related to the juiciness and density of the roasted product. It is particularly suitable for evaluating the effects of myofibrillar protein denaturation and gelation degree caused by high temperature on texture, and provides a mechanical behavior basis for optimizing the roasting process.

[0110] Step S240 , obtaining the elasticity detection value of the meat sample after Chinese-style roasting through texture profile analysis.

[0111] Elasticity describes the ability of meat to return to its original shape after being compressed, and is the same core indicator of TPA as cohesion. In the same compression experiment (conducted simultaneously with S230), the elasticity value is calculated as the ratio of the deformation recovery time of the second compression to the deformation retention time of the first compression (dimensionless). Its principle is based on the elastic modulus of meat gel: during the roasting process, myosin and other proteins are thermally denatured to form a gel network, and its cross-linking strength determines the meat's rebound efficiency. A high elasticity value indicates that the meat has good flexibility and chewing rebound, while a low value reflects tissue hardening or excessive degradation. This parameter is crucial for analyzing the effect of roasting temperature-time combination on protein denaturation, especially for identifying excessive protein aggregation (elasticity reduction) or moderate gelation (elasticity optimization) caused by overheating. It is a direct mechanical indicator for evaluating the effect of roasting process on taste and toughness.

[0112] The meat sample of the present invention may include eye meat, top meat, shoulder meat, shank, rump meat, tenderloin and / or brisket of the sample beef, and the type of the sample cattle may be beef or yak meat.

[0113] The meat samples selected in the examples of the present invention cover key representative parts of beef cattle, including eye meat, top of the head, shoulder meat, shank, rump, shank, and brisket. This multi-site sampling strategy fully reflects the inherent anatomical and functional heterogeneity of the animal muscle system and its inherent influencing mechanism of thermal processing response. Muscles in different parts of the animal perform significantly different biomechanical functions in living animals, directly shaping the unique properties of their microstructure and biochemical composition. For example, the eye meat and top of the head, which are located on the dorsal side of the spine and have less activity, are mainly composed of slow-twitch oxidative muscle fibers that maintain posture. The muscle bundles are slender and the connective tissue wrapping is relatively sparse. At the same time, they have a high potential for intramuscular fat deposition. On the other hand, parts that bear high-load exercise or structural support functions, such as the tendons responsible for limb traction and the brisket that supports the chest and abdomen, are rich in fatigue-resistant fast-twitch glycolytic muscle fibers. Their muscle bundles are wrapped in a thick and tough epimysium. The internal collagen network structure is dense and highly cross-linked, and the lipid deposition space is relatively limited. The shoulder, rump, and butt are in an intermediate state, participating in directional movement, with mixed muscle fiber types, connective tissue diaphragms with thickness and collagen cross-link density between the first two.

[0114] By simultaneously monitoring the changes in the physicochemical indices of these gradient sites before and after grilling (such as collagen thermal solubility, insoluble residue, cohesion, elasticity, etc.), a quantitative relationship map between muscle functional properties and thermal processing adaptability can be constructed. For example, it can be observed that high-motion sites (such as tendons) often require stronger grilling energy input to trigger significant collagen dissolution due to their inherent highly cross-linked collagen structure, but exceeding the critical point may lead to excessive protein aggregation and a sudden drop in elasticity; on the contrary, low-motion sites (such as eye meat) have better initial tenderness, but the rapid melting of intramuscular fat during grilling may affect the cohesion of their internal structure. Moderately loaded sites such as shoulder meat, buttocks, and tenderloins may exhibit the optimal protein gelation state under specific grilling parameter combinations due to their mixed muscle fiber types and moderately cross-linked collagen structure, which is reflected in the peak elasticity. The dense sarcolemmal collagen network in structural support areas such as breast meat requires a specific heat exposure time (such as extending the medium temperature stage) to promote the breaking of its hydrolysis cross-links. Otherwise, a high proportion of insoluble collagen will easily remain, resulting in a tough taste.

[0115] Accordingly, the second aspect of an embodiment of the present invention provides an electronic device comprising: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the above-mentioned method for calculating the index of beef and yak meat suitable for Chinese-style roasting processing.

[0116] Accordingly, a third aspect of an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implements the above-mentioned method for calculating the index of beef and yak meat suitable for Chinese-style roasting.

[0117] The embodiment of the present invention aims to protect a method for calculating an index of beef and yak meat suitable for Chinese-style roasting, comprising the following steps: obtaining first physicochemical property index detection values ​​of meat samples from different parts of several sample cattle before Chinese-style roasting, the first physicochemical property index includes: fat content and raw meat color value; after Chinese-style roasting of meat samples from different parts of several sample cattle according to preset cooking conditions, obtaining second physicochemical property index detection values ​​of each meat sample after Chinese-style roasting, the second physicochemical property index includes: thermal solubility of roasted meat collagen, content of insoluble collagen in roasted meat, cohesiveness and elasticity; based on the first physicochemical property index detection value and the second physicochemical property index detection value, calculating the comprehensive quality index value of each meat sample suitable for Chinese-style roasting cooking method. The above technical solution has the following effects:

[0118] 1. By integrating pre-grilled raw material characteristics (fat content, color) with post-grilled texture indicators (collagen thermal solubility, insoluble collagen content, cohesiveness, and elasticity), a comprehensive quality index across processing stages is constructed. This method breaks through the limitations of traditional methods that analyze single-stage indicators in isolation, quantifies the role of fat melting in promoting heat transfer efficiency and the decisive influence of collagen thermal conversion on tenderness, and achieves closed-loop modeling of the path of grilled quality formation. This model significantly improves the adaptability prediction accuracy of highly variable parts (such as yak shank and tendon meat);

[0119] 2. A purified index set is constructed based on three-level joint screening (Pearson correlation to eliminate redundancy, coefficient of variation to retain sensitivity, and Spearman rank correlation to extract independence). Combined with dual-path fusion weight calculation (principal component analysis to extract high-contribution factors and the CRITIC method to quantify index conflicts), the final weight set is dynamically generated. This method effectively resolves the inherent conflicts in indicators used in traditional static weighting (such as the antagonistic effect of high fat content and low collagen solubility), enabling the comprehensive index to more accurately reflect complex physical and chemical interaction mechanisms.

[0120] 3. Data comparability is achieved through a standardized chain of testing methods. First, before roasting, Soxhlet extraction / near-infrared spectroscopy is used to measure fat and a spectrophotometer is used to determine color. Second, after roasting, collagen solubility is measured by thermal extraction-colorimetry, insoluble collagen is measured by acid hydrolysis and centrifugation, and cohesion / elasticity is determined by texture profile analysis. Third, systematic sampling of multiple parts (7 categories, including eye meat and shoulder meat) and normalization are combined. This standardized framework ensures traceability of data from raw material characteristics to final texture, eliminates interference from human experience (such as subjective bias in traditional roasting trials), and establishes a universal evaluation benchmark for different varieties / parts of beef / yak beef.

[0121] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1A step that specifies a function in one or more boxes.

[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for calculating the index of beef and yak meat suitable for Chinese-style roasting, characterized in that: The steps include: Obtaining first physical and chemical property index test values ​​of meat samples from different parts of several sample cattle before Chinese-style grilling, wherein the first physical and chemical property indexes include fat content and raw meat color value; After Chinese-style grilling of several sample cattle meat samples from different parts according to preset cooking conditions, a second physical and chemical property index test value of each meat sample after Chinese-style grilling is obtained, wherein the second physical and chemical property index includes: thermal solubility of grilled meat collagen, content of insoluble collagen in grilled meat, cohesion and elasticity; Based on the detection values ​​of the first physicochemical property index and the second physicochemical property index, a comprehensive quality index value of each meat sample suitable for Chinese-style grilling cooking method is calculated.

2. The method for calculating the index of beef and yak meat suitable for Chinese-style roasting according to claim 1, characterized in that: Calculating the comprehensive quality index value of each meat sample suitable for Chinese-style grilling cooking based on the first physical and chemical property index detection value and the second physical and chemical property index detection value includes: Normalizing the detection values ​​of each of the first physicochemical property indicators and each of the second physicochemical property indicators, respectively, and obtaining purification index sets for each of the first physicochemical property indicators and each of the second physicochemical property indicators through three-level joint screening; Performing a dual-path fusion calculation on the purified indicator set to obtain a final weight set of the purified indicator set, wherein the dual paths include: a principal component analysis path for extracting initial weights of factor loads of principal components with high contribution rates, and a variation correlation path for calculating indicator conflict coefficients; Performing a weighted sum calculation on the normalized data of the purification index set and the final weight set to obtain a comprehensive quality index value of each meat sample; The calculation formula of the comprehensive quality index value IQI is: Among them, n is the number of indicators contained in the purification indicator set, ω i is the final weight of the i-th indicator, N i is the normalized value of the i-th indicator detection value.

3. The method for calculating the index of beef and yak meat suitable for Chinese-style roasting according to claim 2, characterized in that: The purification index set of each of the first physical and chemical property indexes and each of the second physical and chemical property indexes obtained by the three-level joint screening includes: Calculating a Pearson correlation coefficient matrix of each of the first physicochemical property indices and each of the second physicochemical property indices based on the normalized detection values ​​of each of the first physicochemical property indices and each of the second physicochemical property indices, and eliminating redundant indices with correlations higher than a preset threshold in the Pearson correlation coefficient matrix to obtain a preliminary screening indicator set; Based on the preliminary screening index set, calculating the coefficient of variation of each of the first physicochemical property indexes and each of the second physicochemical property indexes, eliminating low-sensitivity indexes whose coefficient of variation is lower than a preset coefficient of variation threshold, and obtaining an intermediate screening index set; Based on the intermediate screening indicator set, the Spearman rank correlation coefficient of every two indicators is calculated and the Spearman rank correlation coefficient matrix of the intermediate screening indicator set is constructed, and independent indicators whose correlation with other indicators is lower than a preset correlation threshold are retained to obtain a purified indicator set.

4. The method for calculating the index of beef and yak meat suitable for Chinese-style roasting according to claim 3, characterized in that: The performing of dual-path fusion calculation on the purification index set to obtain a final weight set of the purification index set includes: Based on the purification index set, principal component analysis processing is performed, including performing eigenvalue decomposition on the purification index set matrix, extracting principal components whose cumulative variance contribution rate reaches a preset standard, determining the factor loading of each of the first physical and chemical property indicators and each of the second physical and chemical property indicators through eigenvectors, and obtaining an initial weight set of principal components; Based on the purification index set, calculating the coefficient of variation of each of the first physical and chemical property indicators and each of the second physical and chemical property indicators and the rank correlation coefficient with other indicators, and calculating the indicator conflict coefficient using the CRITIC weight method to obtain a conflict weight set; Based on the initial weight set of the principal components and the conflicting weight set, weighted fusion and normalization processing are performed to generate a final weight set of the purification index set.

5. The method for calculating the index of beef and yak meat suitable for Chinese-style roasting according to claim 1, characterized in that: The first physical and chemical property index test values ​​of the meat samples of different parts of several sample cattle before Chinese-style roasting are obtained respectively, including: S1.1: Obtaining a fat content value of the meat sample before Chinese-style grilling by Soxhlet extraction and / or near-infrared spectroscopy; S1.2: Obtain the raw meat color value of the meat sample before Chinese-style grilling by spectrophotometry.

6. The method for calculating the index of beef and yak meat suitable for Chinese-style roasting according to claim 1, characterized in that: The obtaining of the second physical and chemical property index test value of each meat sample after Chinese-style roasting includes: Obtaining the thermal solubility test value of collagen in the meat sample after Chinese-style roasting by thermal extraction-colorimetry; Obtaining the insoluble collagen content of the meat sample after Chinese-style roasting by acid hydrolysis and centrifugation; Obtaining a cohesiveness test value of the meat sample after Chinese-style roasting by texture profile analysis; The elasticity detection value of the meat sample after Chinese-style roasting is obtained by texture profile analysis.

7. The method for calculating the index of beef or yak meat suitable for Chinese-style roasting according to claim 5 or 6, characterized in that: The meat sample includes eye meat, top meat, shoulder meat, shank, rump meat, tenderloin and / or breast meat of the sample beef.

8. The method for calculating the index of beef and yak meat suitable for Chinese-style roasting according to claim 7, characterized in that: The sample cattle include beef cattle and yaks.

9. An electronic device, characterized in that: include: at least one processor; And a memory connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor executes the index calculation method for the suitability of beef and yak meat for Chinese-style roasting processing as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that Computer instructions are stored thereon, and when the instructions are executed by a processor, an index calculation method for the suitability of beef or yak meat for Chinese-style roasting processing as described in any one of claims 1-8 is implemented.

Images