Chinese resident food consumption carbon emission classification method and system based on fuzzy clustering

Abstract

The invention provides a Chinese resident food consumption carbon emission classification method and system based on fuzzy clustering, and belongs to the field of data analysis and application statistical analysis. The invention aims to solve the problem that the existing K-means clustering algorithm depends on the load distance of each region on each principal component factor to carry out region division, is restrained by the fact that the region belongs to a certain index, and is low in practical significance. The method comprises the following steps: firstly, establishing a related index system, performing dimension reduction processing by utilizing principal component analysis, introducing an improved fuzzy C-means algorithm, taking a principal component factor score after dimension reduction as original data of the improved fuzzy C-means algorithm, and introducing a fuzzy clustering effectiveness index to obtain a fuzzy clustering result; the maximum intra-class similarity, the maximum inter-class dissimilarity and the minimum inter-class overlapping degree can be met to the maximum extent, and the optimal classification is determined according to the effectiveness index. The degree of influence of different indexes on each region is well displayed through principal component analysis, various clustering parameters are fully considered, and optimal principle classification is carried out.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis and applied statistical analysis, and in particular to a classification method and system for carbon emissions from food consumption of Chinese residents based on fuzzy clustering. Background Art

[0002] The problem of carbon emissions from food consumption is becoming increasingly serious, and the imbalance of carbon emissions from food consumption between provinces and cities in my country is becoming more and more obvious. In order to study the similarities and differences in carbon emissions from food consumption between regions in my country, and then clarify the reasons for the differences in carbon emissions from food consumption in various regions, it is of great theoretical value and practical significance to study the impact of residents' food consumption on carbon emissions and factor analysis under regional differences.

[0003] In the study of regional differences in carbon emissions from food consumption, the gap in carbon emissions from food consumption in various regions of my country has become increasingly obvious and has caused significant regional imbalances. If this imbalance persists, it will seriously affect my country's overall economic development and ecological balance. To solve this imbalance, we must clearly recognize the similarities and differences in carbon emissions from food consumption between provinces and cities in my country, which involves regional classification. At present, the traditional K-means clustering algorithm can be used to calculate the squared distance between the sample point and the cluster center and iterate the final convergence result, i.e., the cluster center. However, the K-Means algorithm only relies on the load distance of each region on each principal component factor for regional division, and is restricted to belonging to a certain indicator. However, for the data of each province, it is impossible to measure it with one standard, and its practical significance is low. Summary of the invention

[0004] The technical problems to be solved by the present invention are:

[0005] In order to solve the problem that the existing K-means clustering algorithm relies on the load distance of each region on each principal component factor for regional division, sticks to its attribution to a certain indicator, and has low practical significance.

[0006] The present invention adopts the following technical solutions to solve the above technical problems:

[0007] The present invention provides a classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering, comprising the following steps:

[0008] S100. Collect data on influencing factors related to residents’ main food consumption in each provincial administrative region, standardize the data, and select and divide indicators based on variable characteristics and meanings;

[0009] S200, using the improved fuzzy C-means clustering algorithm to calculate the data obtained in step S100, to obtain the membership matrix U and cluster center v j,Through continuous iteration until the change of cluster center is less than the threshold, the iteration ends, and the final category is determined according to the membership hardening;

[0010] S300. In order to verify the effectiveness of clustering, a fuzzy clustering effectiveness index is introduced to enable it to satisfy the maximization of intra-class similarity, inter-class dissimilarity and minimization of inter-class overlap. When the pre-set effectiveness index takes a minimum value, the corresponding number of clusters is taken as the optimal number of clusters.

[0011] Furthermore, in step S100, food consumption carbon emissions are selected as the primary indicator, and economic development level, per capita disposable income, scientific and technological level, total population, urbanization level, environmental protection level, education level, catering industry operating income, urban unit employees, regional freight volume, total output value of agriculture, forestry, animal husbandry and fishery, domestic waste harmless treatment capacity and food consumer price index are selected as secondary indicators.

[0012] Furthermore, in step S200, it specifically includes:

[0013] S210, in order to minimize the objective function J, the Lagrange multiplier method is used for the objective function while satisfying the constraints, that is,

[0014]

[0015] Among them, u ij is the sample point x i With cluster center v j The membership degree of ; m is the fuzzy index, and m>1; d ij is the sample point x i With cluster center v j The distance, d ij The Euclidean distance is used; c is the number of cluster centers; n is the total number of data; λ is the Lagrange multiplier;

[0016] Find the extreme value of function F and get the membership matrix U and cluster center v j ,Right now,

[0017]

[0018] Among them, a single sample x i The sum of the membership of each cluster is 1; k is the number of categories in the data set; d ik is the distance between the sample point and the cluster center;

[0019] S220. When the change in the cluster center calculated by step S210 is greater than the threshold, if the maximum number of iterations has not been reached, the membership matrix is ​​updated, and the cluster center is updated based on the updated membership matrix. The number of iterations is updated, and the change in the cluster center is recalculated until the change in the cluster center is less than the threshold, and the iteration ends.

[0020] Furthermore, in step S220, it specifically includes:

[0021] S221, select the number of categories c and a suitable m, and initialize the matrix U0 determined by the membership function between random values ​​[0,1];

[0022] S222: Calculate the center value v of the cluster according to step S210 j ;

[0023] S223, calculate the new membership matrix U=[u ij ];

[0024] S224, Comparison U j and U j+1 If the changes of both are less than the threshold, the iteration is stopped and the cluster center V and the membership matrix U are output, otherwise go to step S222 to continue the iteration.

[0025] Furthermore, in step S224, the termination condition of the iteration is:

[0026] U k+1 -U k ≤ε′(4)

[0027] Among them, ε′ is the allowed error of the iteration.

[0028] Furthermore, in step S300, it specifically includes:

[0029] S310, introduced intra-class similarity measurement factor:

[0030] in,

[0031] Among them, ε is a fixed value penalty term; is the center of the dataset; is the class i weight;

[0032] S320, introduced inter-class separation measurement factor:

[0033]

[0034] S330, introduced inter-class separation measurement factor:

[0035]

[0036] in,

[0037] S340, a new effectiveness index is constructed by using the intra-class similarity factor, inter-class separation factor, and inter-class overlap factor defined above:

[0038]

[0039] When the effectiveness index reaches a minimum value, the corresponding number of clusters is taken as the optimal number of clusters.

[0040] The present invention provides a classification system for carbon emissions from food consumption of Chinese residents based on fuzzy clustering. The system has a program module corresponding to the above steps, and executes the steps in the above-mentioned classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering when running.

[0041] The present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is configured to implement the steps of a classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering when called by a processor.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The present invention discloses a classification method and system for carbon emissions of Chinese residents' food consumption based on fuzzy clustering. First, a relevant index system is established, and a dimensionality reduction process is performed using principal component analysis. An improved fuzzy C-means algorithm is introduced, and the principal component factor scores after dimensionality reduction are used as the original data of the improved fuzzy C-means algorithm for algorithm improvement and data processing. Finally, a fuzzy clustering effectiveness index is introduced to maximize the intra-class similarity, inter-class dissimilarity and minimized inter-class overlap, and the best classification is determined according to the effectiveness index.

[0044] The present invention proves through experiments that, compared with the existing K-Means algorithm which only relies on the load distance of each region on each principal component factor for regional division, the present invention better demonstrates the degree to which each region is affected by different indicators through principal component analysis, reflecting that the main influencing factors of carbon emissions from food consumption in the region are economic development level, per capita disposable income, total population, food consumer price index, etc., fully considering various clustering parameters and performing optimal principle classification, the main root causes affecting carbon emissions from residents' food consumption can be found according to the analysis, and targeted strategies can be adopted to effectively reduce carbon emissions from residents' food consumption, and a better regional support system can be established. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering in an embodiment of the present invention;

[0046] Figure 2 It is a principal component analysis scree plot in an embodiment of the present invention;

[0047] Figure 3 It is a clustering pedigree diagram of the K-Means clustering algorithm in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In the description of the present invention, it should be noted that the terms "first", "second", and "third" mentioned in the embodiments of the present invention are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second", and "third" may explicitly or implicitly include one or more of the features.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] Specific implementation plan 1: Combine Figure 1 and Figure 2 As shown, the present invention provides a classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering, comprising the following steps:

[0051] S100, collect data on influencing factors related to residents' main food consumption in each provincial administrative region, standardize the data to deal with the relatively scattered data in the feature vector, select and divide indicators according to the characteristics and meaning of the variables,

[0052] Food consumption carbon emissions can be selected as the primary indicator, and 13 indicators including economic development level (100 million yuan), per capita disposable income (yuan), scientific and technological level (%), total population (10,000 people), urbanization level (%), environmental protection level (%), education level (%), catering industry operating income (100 million yuan), urban unit employees (10,000 people), regional freight volume (10,000 tons), total output value of agriculture, forestry, animal husbandry and fishery (100 million yuan), harmless treatment capacity of domestic waste (tons / day) and food consumer price index (previous year = 100) can be selected as secondary indicators. The above 13 indicators are selected by consulting relevant literature and are limited by data availability;

[0053] S200, using the improved fuzzy C-means clustering algorithm to calculate the data obtained in step S100, and obtain the membership matrix U and cluster center v j, through continuous iteration, when the change of cluster center is greater than the threshold, if the maximum number of iterations has not been reached, the membership matrix is ​​updated, and the cluster center is updated based on the updated membership matrix, the number of iterations is updated, and the change of cluster center is recalculated until the change of cluster center is less than the threshold, the iteration ends, and the final category is determined according to the membership hardening; specifically,

[0054] S210, in order to minimize the objective function J, the Lagrange multiplier method is used for the objective function while satisfying the constraints, that is,

[0055]

[0056] Among them, u ij is the sample point x i With cluster center v j The membership degree of m is the fuzzy index (m>1); d ij is the sample point x i With cluster center v j The distance is generally the Euclidean distance; c is the number of cluster centers; n is the total number of data; λ is the Lagrange multiplier;

[0057] Find the extreme value of function F and finally get the membership matrix U and cluster center v j ,Right now,

[0058]

[0059] Among them, a single sample x i The sum of the membership of each cluster is 1; k is the number of categories in the data set; d ik is the distance between the sample point and the cluster center;

[0060] S220, when the change of the cluster center calculated by step S210 is greater than the threshold, if the maximum number of iterations has not been reached, the membership matrix is ​​updated, and the cluster center is updated based on the updated membership matrix, the number of iterations is updated, and the change of the cluster center is recalculated until the change of the cluster center is less than the threshold, and the iteration ends. If the maximum number of iterations is reached, the operation is also terminated; it should be noted that when the maximum number of iterations is reached, the change of the cluster center is less than the threshold after the update, specifically including,

[0061] S221, selecting the number of categories c and the appropriate fuzzy index m, and initializing the matrix U0 determined by the membership function between random values ​​[0, 1]; wherein the matrix U0 determined by the membership function is initialized and initialized between random values ​​[0, 1];

[0062] S222: Calculate the center value v of the cluster according to step S210 j ;

[0063] S223, calculate the new membership matrix U=[u ij ];

[0064] S224, Comparison U j and U j+1 If the changes of both are less than the threshold, the algorithm stops and outputs the cluster center V and the membership matrix U, otherwise it turns to step S222 to continue iterating. The termination condition of the iteration is,

[0065] U k+1 -U k ≤ε′(4)

[0066] Among them, ε′ is the allowed error of iteration;

[0067] S300, in order to verify the effectiveness of clustering, the fuzzy clustering effectiveness index is introduced to maximize the similarity within the class, the difference between the classes and the minimum overlap between the classes, including:

[0068] S310, introduced intra-class similarity measurement factor:

[0069] in,

[0070] Among them, the fixed value penalty term ε, the center of the data set Class i weight The existence of can suppress the monotonically decreasing trend of this factor, making it robust to a certain extent; the smaller the value of C, the higher the similarity within the class, and the better the clustering effect;

[0071] S320, introduced inter-class separation measurement factor:

[0072]

[0073] The square of the maximum class center distance multiplied by the square of the minimum class center distance is used to reflect the separation between classes. The larger the maximum distance and the smaller the minimum distance, the better the separation between classes. The larger the value of S, the better the separation between classes and the better the clustering effect.

[0074] S330, introduced inter-class separation measurement factor:

[0075]

[0076] in,

[0077] O is the sample data x jIn the overlap between classes i1 and i2, u0 is the overlap threshold and is given as 0.2; if the difference in membership of a sample to two classes is greater, it means that the degree to which it belongs to the two classes is farther, and its influence on the entire overlap factor is smaller; the smaller the value of O, the smaller the overlap, and the better the clustering effect;

[0078] S340, a new effectiveness index is constructed by using the intra-class similarity factor, inter-class separation factor, and inter-class overlap factor defined above:

[0079]

[0080] The smaller the new index I is, the better the classification effect is. That is, when it reaches a minimum value, the corresponding number of clusters is the optimal number of clusters.

[0081] Specific implementation scheme 2: The present invention provides a classification system for carbon emissions from food consumption of Chinese residents based on fuzzy clustering. The system has program modules corresponding to the above steps, and executes the steps in the above-mentioned classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering during operation.

[0082] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0083] Specific implementation scheme three: The present invention provides a computer-readable storage medium, which stores a computer program, and the computer program is configured to implement the steps of a classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering when called by a processor.

[0084] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0085] experiment

[0086] According to the China Statistical Yearbook 2022 of the National Bureau of Statistics, we obtained the data on influencing factors related to residents' main food consumption in China's provincial administrative regions in 2022, and divided the indicators according to the characteristics and meanings of the variables. The detailed division is shown in Table 1 below.

[0087] Table 1 Index details

[0088]

[0089]

[0090] The relationship between the components of principal component analysis and the original variables is as follows, where the principal components are the variables obtained by principal component analysis of existing data, and the original variables refer to the 13 variables in Table 1;

[0091]

[0092] Among them, u ij is the i-th component y i and the jth original variable x j The correlation coefficient between

[0093] The correlation matrix is ​​calculated for the standardized data, and the correlation coefficient matrix of the variable is established as follows:

[0094] R=(r ij ) 13×13 =X′X(11)

[0095] The essence of principal component analysis is to reduce the dimension of data and use principal components that contain more information about the total data to represent the linear combination of the original data. Therefore, before using it, you should first check whether the data can be applied to this analysis method. From the derivation process of the principal component, it can be seen that if the covariance matrix of the variable x is Σ, the principal component y of the variable x is a linear combination with the eigenvectors of Σ as coefficients. They are independent eigenroots with variance Σ, and they are required to meet the application conditions of principal component analysis: λ1≥λ2≥...≥λ 13 >0, and when running the code through RStudio software, it is found that the eigenvalue of the correlation coefficient matrix meets this condition, so the principal component analysis method can be used. The detailed data are shown in Table 2 below.

[0096] Table 2 Correlation coefficient eigenvalues ​​of principal component analysis

[0097]

[0098] According to the main component standard deviation, contribution rate, and cumulative variance contribution rate in Table 3 below, the cumulative variance contribution rate of the first four main variables is 87.86%, so the first four main variables can be used to reflect the original evaluation object; combined with Figure 2 It can also be seen that the trend becomes stable after the turning point at 4, so the number of principal components is more appropriate to be 4;

[0099] Table 3. Principal component standard deviation, contribution rate, cumulative variance contribution rate

[0100]

[0101] Based on the coefficients of each factor on each variable in the principal component factor loading matrix Table 4, the four extracted principal component factors are explained according to the output results:

[0102] (1) The main component Comp.1 is in the economic development level x1, urban unit employment x9, food consumer price index x 13 There is a large load on it, which reflects the driving effect of economic development on carbon emissions from food consumption. y1 is defined as the economic driving factor;

[0103] (2) The principal component Comp.2 has a large load on per capita disposable income x2, total population x4, and urbanization level x5, reflecting the impact of factors related to population consumption capacity on food consumption carbon emissions. y2 is defined as the population effect factor;

[0104] (3) The main component Comp.3 is in the environmental protection level x6, catering industry operating income x8, and total output value of agriculture, forestry, animal husbandry and fishery x 11 , harmless treatment capacity of domestic waste x 12 There is a large load on it, which reflects the level of environmental and food consumption factors that are directly linked to carbon emissions from food consumption, and y3 is defined as the internal development factor;

[0105] (4) The main component Comp.4 is in the science and technology level x3, education level x7, regional freight volume x 10 There is a large load on it, reflecting the indirect impact of input factors in science and technology, education, etc. on carbon emissions from food consumption. y4 is defined as the external environmental factor.

[0106] Therefore, by combining the scores of the first four principal components of each province, city and region, we can obtain the degree of influence of the factor indicators of food consumption carbon emissions in each region.

[0107] Table 4 Factor loading matrix

[0108]

[0109]

[0110] Taking the correlation coefficient of each principal component to the original index as the weight, each principal component is expressed as a linear combination of the original index. From the above analysis, it can be seen that the economic significance of the principal component is determined by the comprehensive consideration of the index with a larger weight in each principal component. Here, the comprehensive significance of carbon emissions from food consumption is used to determine it, that is:

[0111] C j =a j1 x1+a j2 x2+a j3 x3+a j4 x4,j=1,2,3,4(12)

[0112] Among them, a j1 、a j2 、a j3 、a j4 Data parameters calculated for the four principal components;

[0113] According to the scree plot and weight analysis, the variance contribution rate of the first four indicators can be used as the weight to obtain their linear combination and calculate the principal component score, namely:

[0114]

[0115] Using the variance contribution rate calculated previously, we can get:

[0116]

[0117] The final main factor scores and rankings are shown in Table 5.

[0118] Table 5 Main factor scores and rankings

[0119]

[0120]

[0121] From Table 5 above, we can see that Guangdong Province is most affected by economic driving factors, followed by Jiangsu Province, Shandong Province, and Zhejiang Province, and Guangdong Province is significantly higher than other regions, while the lower-ranked ones are Tibet Autonomous Region, Qinghai Province, Ningxia Hui Autonomous Region, Gansu Province and other underdeveloped western regions. That is, under the influence of economic driving factors such as economic development level, number of employees in urban units, and food consumer price index, which are biased towards regional economy and population economy, the carbon emission levels of food consumption in Guangdong Province, Jiangsu Province and other places are much higher than other regions; the provinces with large populations such as Henan Province, Shandong Province, and Sichuan Province are more affected by population factors. In comparison, Beijing, Shanghai, Tianjin and other regions are less affected by this factor.

[0122] The above data are processed by fuzzy clustering, and the 4 main factors extracted are used to replace the original 13 indicators as the initial value of the fuzzy clustering algorithm. In order to avoid excessive errors caused by the sensitivity of the initial clustering center, the FCM algorithm is executed 10 times and the average value is taken when calculating the fuzzy clustering effectiveness index. The clustering situation in 2022 is obtained by running the FCM algorithm, as shown in Table 6 below.

[0123] Table 6 Fuzzy matrix cluster centers when stopping iteration

[0124]

[0125] It can be seen from Table 6 above that when the fuzzy C-means clustering algorithm is run until the iteration stops, the data set is divided into 5 categories. Therefore, it can be determined that the most effective and reasonable number of clusters is 5. Therefore, it is most reasonable to divide China's 31 provincial administrative regions (except Hong Kong, Macao and Taiwan) into 5 categories in the present invention; by observing the output membership matrix, each sample is attributed to the category where the maximum membership value is located, and the final classification of each region is determined. The results are shown in Table 7 below:

[0126] Table 7 Regional division under fuzzy C-means clustering algorithm

[0127]

[0128] In order to compare the fuzzy C-means clustering algorithm with the general clustering algorithm (taking the K-Means algorithm as an example), the scores of the four principal component factors of China's 31 provincial administrative regions (excluding Hong Kong, Macao and Taiwan) in 2022 were used as the original data of the K-Means algorithm for classification, and the following clustering pedigree diagram and regional division details were obtained. The specific results are as follows Figure 3 shown.

[0129] Through analysis, it can be seen that the number of regional classification divisions of the two is different. The improved fuzzy C-means clustering algorithm of the present invention divides the regions into 5 categories, while the K-Means algorithm divides them into 6 categories. The main reason is that the improved fuzzy C-means clustering algorithm of the present invention continuously iterates according to the pre-set effectiveness index and finally iterates the optimal number of clusters, and divides the regions according to the membership of each region in different clusters; the K-Means algorithm continuously calculates the squared distance between the sample point and the cluster center and iterates the final convergence result, i.e., the cluster center, so the two clustering algorithms have different numbers of clusters.

[0130] Secondly, the improved fuzzy C-means clustering algorithm of the present invention focuses on classifying regions with the same degree of influence of factors and positive and negative effects into one category, and each clustering region has a clear category division: Beijing, Tianjin, Shanghai, and Zhejiang in the first category are all highly developed regions that are greatly affected by economic factors and external environmental factors. Environmental improvements in food consumption carbon emissions in these regions must focus on the economic aspect, develop a green economy and a low-carbon economy, which will not hinder the original economic development of the region, and can minimize the adverse impact of the economy on the ecological environment; Anhui, Fujian, Jiangxi, Shandong, Guangxi, Hainan, and Guizhou in the second category are mainly affected by the external environment and population. The more population in a region, the easier it is to cause food consumption carbon emission pollution, and more investment in environmental funds for regional improvement will reduce environmental pressure. The dual influence of positive and negative effects classifies the above-mentioned regions into one category. This type of region should reduce regional population inflow as much as possible on the basis of strengthening environmental protection investment to ensure that other impacts are minimized, and reduce the regional population load deficit; Jiangsu, Henan, Hubei, Hunan, Guangdong, Sichuan, Yunnan, and Xinjiang in the third category Xinjiang is mainly affected by the dual factors of human factors and economic development. Both the economy and population have a negative impact on the ecological environment of food consumption. Therefore, the environmental improvement of carbon emissions from food consumption in this type of region should focus on improving the economic development model and reducing the regional population load deficit; Hebei, Shanxi, Chongqing, Shaanxi, Gansu, Qinghai, and Ningxia in the fourth category are mainly affected by external environmental and economic factors. The environmental improvement of carbon emissions from food consumption in this type of region should focus on developing a green economy in accordance with local conditions, so that the geographical environment can be transformed from a shortcoming to an economic development advantage in accordance with local conditions, thereby reducing the impact of regional carbon emissions from food consumption on the environment; Inner Mongolia, Liaoning, Jilin, Heilongjiang, and Tibet in the fifth category are composed of the three northeastern provinces and the two regions with the largest carbon emissions from food consumption. The restrictions of geographical factors and regional eating habits have classified these five regions into one category. The environmental improvement of carbon emissions from food consumption in this type of region should focus on improving residents' dietary awareness, enhancing the awareness of dietary diversity, reducing the consumption of animal food as much as possible, and advocating a diversified diet suitable for local advantages. In contrast, the K-Means algorithm only relies on the load distance of each region on each principal component factor to divide the region, but its practical significance is relatively low.

[0131] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A classification method for carbon emissions from food consumption of Chinese residents based on fuzzy clustering, characterized in that: The following steps are involved: S100. Collect data on influencing factors related to residents’ main food consumption in each provincial administrative region, standardize the data, and select and divide indicators based on variable characteristics and meanings; S200, using the improved fuzzy C-means clustering algorithm to calculate the data obtained in step S100, and obtain the membership matrix U and cluster center v j ,Through continuous iteration until the change of cluster center is less than the threshold, the iteration ends, and the final category is determined according to the membership hardening; S300. In order to verify the effectiveness of clustering, a fuzzy clustering effectiveness index is introduced to enable it to satisfy the maximization of intra-class similarity, inter-class dissimilarity and minimization of inter-class overlap. When the pre-set effectiveness index takes a minimum value, the corresponding number of clusters is taken as the optimal number of clusters.

2. The classification method of carbon emissions from food consumption of Chinese residents based on fuzzy clustering according to claim 1 is characterized by: In step S100, food consumption carbon emissions are selected as the primary indicator, and economic development level, per capita disposable income, scientific and technological level, total population, urbanization level, environmental protection level, education level, catering industry operating income, urban unit employees, regional freight volume, total output value of agriculture, forestry, animal husbandry and fishery, domestic waste harmless treatment capacity and food consumer price index are selected as secondary indicators.

3. According to claim 1, a classification method for carbon emissions from Chinese residents' food consumption based on fuzzy clustering is characterized in that: In step S200, it specifically includes: S210, in order to minimize the objective function J, the Lagrange multiplier method is used for the objective function while satisfying the constraints, that is, Among them, u ij is the sample point x i With cluster center v j The membership degree of ; m is the fuzzy index, and m>1; d ij is the sample point x i With cluster center v j The distance, d ij The Euclidean distance is used; c is the number of cluster centers; n is the total number of data; λ is the Lagrange multiplier; Find the extreme value of function F and get the membership matrix U and cluster center v j ,Right now, Among them, a single sample x i The sum of the membership of each cluster is 1; k is the number of categories in the data set; d ik is the distance between the sample point and the cluster center; S220. When the change in the cluster center calculated by step S210 is greater than the threshold, if the maximum number of iterations has not been reached, the membership matrix is ​​updated, and the cluster center is updated based on the updated membership matrix. The number of iterations is updated, and the change in the cluster center is recalculated until the change in the cluster center is less than the threshold, and the iteration ends.

4. According to claim 3, a classification method for carbon emissions from Chinese residents' food consumption based on fuzzy clustering is characterized in that: In step S220, it specifically includes: S221, select the number of categories c and a suitable m, and initialize the matrix U0 determined by the membership function between random values ​​[0,1]; S222: Calculate the center value v of the cluster according to step S210 j ; S223, calculate the new membership matrix U=[u ij ]; S224, Comparison U j and U j+1 If the changes of both are less than the threshold, the iteration is stopped and the cluster center V and the membership matrix U are output, otherwise go to step S222 to continue the iteration.

5. According to claim 4, a classification method for carbon emissions from Chinese residents' food consumption based on fuzzy clustering is characterized in that: In step S224, the termination condition of the iteration is: ||In k+1 -IN k ||≤ε′ (4) Among them, ε′ is the allowed error of the iteration.

6. According to claim 4, a classification method for carbon emissions from Chinese residents' food consumption based on fuzzy clustering is characterized in that: In step S300, it specifically includes: S310, introduced intra-class similarity measurement factor: in, Among them, ε is a fixed value penalty term; is the center of the dataset; is the class i weight; S320, introduced inter-class separation measurement factor: S330, introduced inter-class separation measurement factor: in, S340, a new effectiveness index is constructed by using the intra-class similarity factor, inter-class separation factor, and inter-class overlap factor defined above: When the effectiveness index reaches a minimum value, the corresponding number of clusters is taken as the optimal number of clusters.

7. A classification system for carbon emissions from food consumption of Chinese residents based on fuzzy clustering, characterized by: The system has a program module corresponding to the steps of any one of claims 1 to 6 above, and executes the steps in the above-mentioned classification method of carbon emissions from food consumption of Chinese residents based on fuzzy clustering when running.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the method for classifying carbon emissions from food consumption of Chinese residents based on fuzzy clustering according to any one of claims 1 to 6 when called by a processor.

Images