Static method and device for regional classification of implicit carbon emission in multi-regional environment and storage medium
By constructing a multi-type objective function model and cluster analysis, we can identify the role characteristics of multiple regions in trade implicit carbon emissions, solve the problem of insufficient allocation of carbon emission reduction targets in the existing technology, and achieve more accurate regional carbon emissions Responsibility division and carbon emission accounting strategies.
Patent Information
- Application Number
- CN202310497706.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-05
- Publication Date
- 2024-11-08
AI Technical Summary
The existing technology cannot effectively identify the role characteristics of the region in trade implicit carbon emissions in multi-regional carbon emission accounting, resulting in insufficient allocation of carbon emission reduction targets and lack of in-depth analysis, especially the in-depth characteristics that cannot be extracted based on traditional statistical methods. .
Build a multi-type objective function model, including the flow objective function model of implicit carbon emissions and the value-based trade flow objective function model, and identify the role characteristics of the region in the implicit carbon emissions of trade through cluster analysis, and use machine learning algorithms such as K-means The clustering algorithm extracts the depth features of the data.
Improve the accuracy of regional carbon emission responsibilities, and provide a new carbon emission accounting strategy, which can examine trade flows and implicit carbon emission flows in multiple regions or countries under static competition, overcoming the lack of based on the existing technology. Issues considered in the 'consumer responsibility system' starting from the carbon emission flow network.
Smart Images

Figure CN118916799A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to a method for classifying carbon emission regions, which belongs to the field of carbon emission accounting, and in particular to a static method, device and storage medium for regional classification of implicit carbon emissions in a multi-regional environment. Background Art
[0002] The inventory compilation method from the IPCC National Greenhouse Gas Inventory Guidelines is widely used internationally to calculate greenhouse gases produced within the national territory, that is, the producer responsibility system. However, in order to meet their consumption needs, developed countries reduce carbon emissions by importing carbon-intensive products from developing countries and transferring high-carbon-intensive industries. The above accounting method only confirms carbon emission responsibilities from the perspective of direct production, which is not conducive to net carbon exporting countries. There are similar carbon reduction target allocation issues for domestic administrative regions or industries.
[0003] At present, the technical characteristics of each region's environmental carbon emission roles are not obvious, the basis for the division of regional carbon emission responsibilities is not sufficient, and there is a lack of in-depth analysis of the value flow and carbon emission environmental emission role allocation technology in multiple regions across the country under medium and macro scenarios. In particular, traditional statistical methods cannot extract the deep characteristics of the data and can only process data under linear infinite approximation. Summary of the invention
[0004] According to the first aspect of the present application, a static method for regional classification of multi-regional environmental embodied carbon emissions is provided. The method can conduct in-depth analysis of the carbon emission data of each region and identify the role characteristics of each region in trade embodied carbon emissions, so as to serve as a reasonable basis for the division of regional carbon emission responsibilities.
[0005] The static method for regional classification of embodied carbon emissions in multiple regions includes:
[0006] Aiming at the implicit carbon emissions in multiple regions, a multi-type objective function model is constructed, and deep features are extracted from the function data of the multi-type objective function model to obtain regional classification results.
[0007] Preferably, the multi-type objective function model includes an implicit carbon emission flow objective function model and a value-based trade flow objective function model, and cluster analysis is performed on the function data of the two models; based on the cluster analysis results, the role characteristics of multiple regions in trade implicit carbon emissions are identified.
[0008] Preferably, the method further comprises: statically displaying the region classification result.
[0009] Preferably, the implicit carbon emission flow objective function model is as shown in the following formula:
[0010]
[0011] Among them, ΔG k represents the net carbon emissions from the kth region to other regions, represents the total implicit carbon emissions of the kth region from the perspective of consumption, represents the total direct production carbon emissions of the kth region from the production perspective, n represents the number of regions within the measured classification range; ΔG k A positive or negative value indicates that the environmental carbon emissions generated by regional consumption are more or less than those generated by production.
[0012] Preferably, the value-based trade flow objective function model is as shown in the following formula:
[0013]
[0014] Among them, ΔM k represents the net value flow from the kth region to other regions, represents the total consumption of the kth region, represents the total output of the kth region, n represents the number of regions included in the research scope; ΔM k A positive or negative value indicates that the consumption value of the region is more or less than the production value.
[0015] Preferably, the cluster analysis of the function data of the two models includes: clustering the net carbon emission quantity ΔG from the kth region to other regions k and the net value flow ΔM k The data is input into a clustering algorithm, and a region type is obtained through the clustering algorithm, where the region type is a consumption region or a production region; the clustering algorithm includes at least one of the following: a K-means clustering algorithm and a K-means+ clustering algorithm.
[0016] Preferably, the method further comprises: before constructing the implicit carbon emission flow objective function model, calculating the carbon emission flow matrix, and performing the following steps:
[0017] Assume that the carbon emission and production markets in each region are perfectly competitive markets;
[0018] Obtain value-based input-output models and direct carbon emission data information;
[0019] Determine the direct consumption coefficient of each region according to the value-based input-output model and calculate the Leontief inverse matrix;
[0020] The direct carbon emission data information is represented as a diagonal matrix, and a point multiplication is performed between the diagonal matrix and the inverse Leontief matrix to obtain a carbon emission flow matrix.
[0021] Preferably, the carbon emission flow matrix is expressed as:
[0022] G=F diag (IA) -1 Y diag
[0023] Among them, F diag is the diagonal matrix of direct carbon flow generated by carbon emissions per unit of gross output value of each region in the tested classification, that is, carbon emission density; A represents the direct consumption coefficient matrix of the input-output model; I is the unit matrix; Y represents the final use column vector of each region, G is the carbon emission flow matrix, and diag represents the corresponding diagonal matrix.
[0024] According to a second aspect of the present application, a static device for regional classification of embodied carbon emissions in a multi-region environment is provided, comprising:
[0025] The calculation module is used to build multi-type objective function models for multi-regional embodied carbon emissions;
[0026] The classification module is used to extract deep features of the function data of the multi-type objective function model to obtain a regional classification result.
[0027] Preferably, the device further comprises a display module for statically displaying the region classification result.
[0028] Preferably, the multi-type objective function model constructed by the calculation module includes an implicit carbon emission flow objective function model and a value-based trade flow objective function model; the calculation module is used to perform cluster analysis on the function data of the two models, and identify the role characteristics of multiple regions in trade implicit carbon emissions based on the clustering results.
[0029] Preferably, the implicit carbon emission flow objective function model is as shown in the following formula:
[0030]
[0031] Among them, ΔG k represents the net carbon emissions from the kth region to other regions, represents the total implicit carbon emissions of the kth region from the perspective of consumption, represents the total direct production carbon emissions of the kth region from the production perspective, n represents the number of regions within the measured classification range; ΔG k A positive or negative value indicates that the environmental carbon emissions generated by regional consumption are more or less than those generated by production.
[0032] Preferably, the value-based trade flow objective function model is as shown in the following formula:
[0033]
[0034] Among them, ΔM k represents the net value flow from the kth region to other regions, represents the total consumption of the kth region, represents the total output of the kth region, n represents the number of regions included in the research scope; ΔM k A positive or negative value indicates that the consumption value of the region is more or less than the production value.
[0035] According to a third aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, some or all of the steps in the above-mentioned static method for regional classification of embodied carbon emissions in a multi-regional environment are implemented.
[0036] The beneficial effects of this application include:
[0037] 1) The implicit carbon emission flow objective function model provided in this application, in addition to the input-output model, also incorporates carbon emission flow data as a new variable into the model during construction, examines the trade flows and implicit carbon emission flows in multiple regions of a country or multiple regions of multiple countries under static competition, and provides a new carbon emission accounting strategy.
[0038] 2) This application adopts machine learning algorithms, such as K-means clustering algorithm or K-means+ clustering algorithm, which can extract deep features of carbon emission flow and value flow data, solving the technical problem that the existing technology can only use statistical methods but cannot extract deep features of data; using deep features as an identifier of the value role allocation of different regions or provinces in the carbon emission environment improves the accuracy of regional carbon emission responsibility division, and overcomes the technical problem that the existing technology lacks the consideration of "consumer responsibility system" based on the carbon emission flow network in the research of multi-regional carbon emission reduction analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 A schematic diagram of the application process of the static method for regional classification of embodied carbon emissions in multi-regional environments in one embodiment of the present application in provincial classification;
[0040] Figure 2 A schematic diagram of the relationship between the value-based input-output model and direct carbon emission data information in one implementation of the present application;
[0041] Figure 3 This is a schematic diagram of the environmental implicit carbon emission network in one embodiment of the present application;
[0042] Figure 4This is the clustering result of the allocation of roles for environmental implicit carbon emissions in different provinces in one embodiment of the present application. DETAILED DESCRIPTION
[0043] The present application is described in detail below with reference to embodiments, but the present application is not limited to these embodiments.
[0044] As an example of this application, see Figure 1 , which shows a flow chart of a static method for regional classification of multi-regional environmental embodied carbon emissions, which is applied to provincial regional carbon emission regional classification, and the method includes:
[0045] S1: Obtain the value-based input-output model and direct carbon emission data information related to the clustering.
[0046] S11: The tested categories are determined to be 30 provinces across the country (due to the lack of basic data, Tibet, Hong Kong and Macao are not within the scope of the study).
[0047] S12: Assume that the domestic carbon emission and production markets are perfectly competitive markets; the products or services obtained by each region from other regions are completely used in the region, and the production inputs within the region all come from regional resources and serve the final use of the region. Therefore, the implicit carbon emissions caused by trade flows in various domestic regions also come only from regional emissions.
[0048] S13: Direct consumption coefficient determined based on the constructed annual national multi-regional input-output model.
[0049] S14: Set the total annual carbon dioxide emissions of each province as an exogenous variable and collect data.
[0050] S2: According to the environmental emission coefficient, the implicit carbon emission flow matrix and the value-based trade flow matrix are constructed; wherein the implicit carbon emission flow matrix is composed of As matrix elements, represents the total implicit carbon emissions of province k from the perspective of consumption; the value-based trade flow matrix is As matrix elements, Represents the total consumption of province or region k.
[0051] S21: Calculate the carbon emission flow matrix G according to the environmental emission coefficient:
[0052] G=F diag (IA) -1 Y diag
[0053] Among them, F diagContains environmental emission coefficients, which is a diagonal matrix of direct carbon flow generated by carbon emissions per unit of gross output value in the region, that is, carbon emission density. A represents the direct consumption coefficient matrix of the input-output model, which is an input-output coefficient matrix reflecting the technical level. I is the unit matrix. Y represents the final use column vector of each region.
[0054] S3: Based on the above carbon emission flow matrix G, an implicit carbon emission flow objective function model is constructed; according to the value-based input-output model, a value-based trade flow objective function model is constructed to calculate the net carbon emissions and net trade from province k to other provinces or regions respectively.
[0055] S31: The implicit carbon emission flow objective function model is shown in the following formula:
[0056]
[0057] Among them, ΔG k represents the net carbon emissions from province k to other provinces or regions, represents the total implicit carbon emissions of province k from the perspective of consumption, represents the total direct production carbon emissions of province k from the production perspective, n represents the number of regions within the measured classification range; ΔG k A positive or negative value indicates that the environmental carbon emissions generated by consumption in the province are more or less than those generated by production.
[0058] S32: The value-based trade flow objective function model is shown in the following formula:
[0059]
[0060] Among them, ΔM k represents the inter-provincial or inter-regional net value flow from province k to other regions, represents the total consumption of province or region k, represents the total output of k provinces or regions, n represents the number of regions included in the research scope; ΔM k A positive or negative value indicates that the province's consumption value is more or less than its production value.
[0061] S4: Based on machine learning clustering algorithm, ΔG k and ΔM k Perform cluster analysis, identify provinces with "net consumption" and "net production" categories of trade-implied carbon emissions based on the clustering results, and identify the role characteristics of the region or province in carbon emissions. The machine learning clustering algorithm includes but is not limited to the K-means clustering algorithm, or the K-means+ clustering algorithm. The machine learning clustering process can be programmed using Python language or matlab.
[0062] The world regional part of the model is similar to the one-country-multiple-regions case.
[0063] like Figure 2 As shown, the relationship between the value-based input-output model and the direct carbon emission data information in one embodiment of the present application is: A represents the direct consumption coefficient matrix of the input-output model, which is an input-output coefficient matrix reflecting the technical level. X and Y represent the total output column vector and the final use column vector of each region. C represents the total annual carbon dioxide emissions column vector of each province.
[0064] Using the environmental emission coefficient, an environmental extended input-output analysis is performed by adding additional columns, such as energy use, carbon emissions, water consumption, and air pollutants. Direct carbon emissions are used as a new column for the environmental emission coefficient of the input-output model, and the carbon emission flow matrix between each regional grid is calculated through a combined equation.
[0065] This embodiment uses China's input-output data in 2017 as an example to conduct research. The data on energy input to each sector and each region can be used to evaluate the carbon emissions reflected in trade between 42 industries or 30 regions.
[0066] Figure 3 This is a schematic diagram of the environmental embodied carbon emission network according to an embodiment of the present invention. From the perspective of the carbon emission flow network based on production and consumption in each province, the embodied carbon in the products consumed by the province itself is the largest, followed by emissions transferred to other regions. This is because the products produced in a region are generally used to meet their own needs first, and the surplus is exported to other regions.
[0067] Figure 4 It is the clustering result of the role allocation of environmental embodied carbon emissions in different provinces according to an embodiment of the present invention.
[0068] K-means algorithm clustering is one of the most popular and elegant clustering methods in machine learning, which is used to divide a data set into K different, non-overlapping clusters. In the context of machine learning, K-means cluster analysis is an unsupervised task, which is a clustering algorithm based on the common characteristics of a sample set. First, the algorithm automatically selects K seeds as the initial centroids of each of the K classes; then calculates the distance between each seed and the centroid of each class, and re-divides all seeds into K classes according to the distance; then recalculates the centroid of each class... The above calculation process is iterated many times until the function converges, that is, the cluster center no longer changes. In the embodiment, K-means algorithm clustering is used to establish a regional evaluation model based on the economic carbon emission difference data in the input-output model. The algorithm clusters the vectors of n = 30 provinces; the sample dimension m = 2 is set, that is, the regional economic difference and carbon emission difference, and it is divided into clusters of k by the K-means algorithm to determine the provinces with high or low value difference / carbon emission difference with common characteristics. By extracting the deep features of carbon emission flow and value flow data, the value roles of different regions or provinces in the carbon emission environment can be accurately identified.
[0069] In the embodiment, the provinces are comprehensively clustered by k-means clustering and its components. The clustering results are evaluated by elbow method and CH method. The typical characteristics of different provincial clustering results are obtained. Programming can be realized by python language or matlab to complete K-means clustering.
[0070] The clustering results of this embodiment are shown in Table 1:
[0071] Table 1
[0072]
[0073]
[0074] Cluster 1 and cluster 3 have obvious role category characteristics. Combined with the data for comparative analysis, Beijing, Zhejiang and Guangdong Province in cluster 1 consume a large number of products and services produced by other provinces through consumption in embodied carbon emissions and regional trade relations. The inter-provincial trade brought about by this demand leads to a large amount of environmental embodied carbon emissions transfer, making the environmental embodied carbon emissions of the three provinces in this cluster much higher than their direct production carbon emissions. Cluster 3 is exactly the opposite of cluster 1. In the flow of environmental embodied carbon emissions, the three typical provinces of cluster 3 - Hebei, Shanxi and Inner Mongolia - sell a large number of products and services produced in the region, thereby realizing the outflow of embodied carbon to other regions. Further analysis shows that although the differences in embodied carbon emissions in various regions and their trade flow differences show a certain convergence, they are not completely consistent.
[0075] The calculation results after multiple iterations of the model indicate that Beijing, Zhejiang and Guangdong provinces in cluster 1 are representatives of economically developed provinces. They act as "buyers" or net "consumers" in the transfer of environmentally embodied carbon emissions, occupy a dominant position in the carbon emission network, and purchase products and services from carbon emission exporting provinces; while Hebei, Shanxi and Inner Mongolia provinces in cluster 3, as "sellers" or net "producers", are relatively disadvantaged net exporting provinces in the environmental carbon emission network.
[0076] This example shows that the direct carbon emissions calculated from the perspective of the production responsibility system in some provinces appear to be small, but if the environmentally implicit carbon emissions transferred from this region to other regions are taken into account, a different conclusion may be drawn.
[0077] The present application also provides a static device for regional classification of embodied carbon emissions in multiple regional environments, comprising:
[0078] The calculation module is used to build multi-type objective function models for multi-regional embodied carbon emissions;
[0079] The classification module is used to extract deep features of the function data of the multi-type objective function model to obtain a regional classification result.
[0080] In one embodiment, the device further includes a display module for statically displaying the region classification result.
[0081] In one embodiment, the multi-type objective function model constructed by the calculation module includes an implicit carbon emission flow objective function model and a value-based trade flow objective function model; the calculation module is used to perform cluster analysis on the function data of the two models, and identify the role characteristics of multiple regions in trade implicit carbon emissions based on the clustering results.
[0082] In one embodiment, the implicit carbon emission flow objective function model is shown in the following formula:
[0083]
[0084] Among them, ΔG k represents the net carbon emissions from the kth region to other regions, represents the total implicit carbon emissions of the kth region from the perspective of consumption, represents the total direct production carbon emissions of the kth region from the production perspective, n represents the number of regions within the measured classification range; ΔG k A positive or negative value indicates that the environmental carbon emissions generated by regional consumption are more or less than those generated by production.
[0085] In one embodiment, the value-based trade flow objective function model is shown in the following formula:
[0086]
[0087] Among them, ΔM k represents the net value flow from the kth region to other regions, represents the total consumption of the kth region, represents the total output of the kth region, n represents the number of regions included in the research scope; ΔM k A positive or negative value indicates that the consumption value of the region is more or less than the production value.
[0088] The present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements part or all of the steps in the above-mentioned static method for regional classification of embodied carbon emissions in a multi-regional environment.
[0089] The above are only a few embodiments of the present application and do not constitute any form of limitation to the present application. Although the present application is disclosed as above with preferred embodiments, it is not intended to limit the present application. Any technician familiar with the profession, without departing from the scope of the technical solution of the present application, using the technical content disclosed above to make slight changes or modifications are equivalent to equivalent implementation cases and fall within the scope of the technical solution.
Claims
1. A static method for regional classification of embodied carbon emissions in multiple regions, characterized in that: The method includes: constructing a multi-type objective function model for multi-region implicit carbon emissions, extracting deep features of function data of the multi-type objective function model, and obtaining regional classification results.
2. The static method for regional classification of embodied carbon emissions in multiple regions according to claim 1, characterized in that: The multi-type objective function model includes an implicit carbon emission flow objective function model and a value-based trade flow objective function model, and cluster analysis is performed on the function data of the two models; based on the cluster analysis results, the role characteristics of multiple regions in trade implicit carbon emissions are identified; Preferably, the method further comprises: statically displaying the region classification result.
3. The static method for regional classification of embodied carbon emissions in multiple regions according to claim 2, characterized in that: The implicit carbon emission flow objective function model is shown in the following formula: Among them, ΔG k represents the net carbon emissions from the kth region to other regions, represents the total implicit carbon emissions of the kth region from the perspective of consumption, represents the total direct production carbon emissions of the kth region from the production perspective, n represents the number of regions within the measured classification range; ΔG k A positive or negative value indicates that the environmental carbon emissions generated by regional consumption are more or less than those generated by production.
4. The static method for regional classification of embodied carbon emissions in multiple regions according to claim 2, characterized in that: The value-based trade flow objective function model is shown in the following formula: Among them, ΔM k represents the net value flow from the kth region to other regions, represents the total consumption of the kth region, represents the total output of the kth region, n represents the number of regions included in the research scope; ΔM k A positive or negative value indicates that the consumption value of the region is more or less than the production value.
5. The static method for regional classification of embodied carbon emissions in multiple regions according to claim 2, characterized in that: The cluster analysis of the function data of the two models includes: clustering the net carbon emission quantity ΔG from the kth region to other regions k and the net value flow ΔM k The data is input into a clustering algorithm, and a region type is obtained through the clustering algorithm, where the region type is a consumption region or a production region; the clustering algorithm includes at least one of the following: a K-means clustering algorithm and a K-means+ clustering algorithm.
6. The static method for regional classification of embodied carbon emissions in multiple regions according to claim 3, characterized in that: Before constructing the implicit carbon emission flow objective function model, the carbon emission flow matrix is calculated, including the following steps: Assume that the carbon emission and production markets in each region are perfectly competitive markets; Obtain value-based input-output models and direct carbon emission data information; Determine the direct consumption coefficient of each region according to the value-based input-output model and calculate the Leontief inverse matrix; The direct carbon emission data information is represented as a diagonal matrix, and a point multiplication is performed between the diagonal matrix and the inverse Leontief matrix to obtain a carbon emission flow matrix.
7. The static method for regional classification of embodied carbon emissions in multiple regions according to claim 6, characterized in that: The carbon emission flow matrix is expressed as: G=F diag (I-A) -1 Y diag Among them, F diag is the diagonal matrix of direct carbon flow generated by carbon emissions per unit of gross output value of each region in the tested classification; A represents the direct consumption coefficient matrix of the input-output model; I is the unit matrix; Y represents the final use column vector of each region, G is the carbon emission flow matrix, and diag represents the corresponding diagonal matrix.
8. A static device for regional classification of embodied carbon emissions in multiple regions, characterized in that: The device comprises: The calculation module is used to build multi-type objective function models for multi-regional embodied carbon emissions; The classification module is used to extract deep features of the function data of the multi-type objective function model to obtain a regional classification result.
9. The static device for regional classification of embodied carbon emissions in multiple regions according to claim 8, characterized in that: The device also includes a display module, which is used to statically display the region classification result; Preferably, the multi-type objective function model constructed by the calculation module includes an implicit carbon emission flow objective function model and a value-based trade flow objective function model; the calculation module is used to perform cluster analysis on the function data of the two models, and identify the role characteristics of multiple regions in trade implicit carbon emissions according to the clustering results; Preferably, the implicit carbon emission flow objective function model is as shown in the following formula: Among them, ΔG k represents the net carbon emissions from the kth region to other regions, represents the total implicit carbon emissions of the kth region from the perspective of consumption, represents the total direct production carbon emissions of the kth region from the production perspective, n represents the number of regions within the measured classification range; ΔG k A positive or negative value indicates that the environmental carbon emissions generated by regional consumption are more or less than those generated by production; Preferably, the value-based trade flow objective function model is as shown in the following formula: Among them, ΔM k represents the net value flow from the kth region to other regions, represents the total consumption of the kth region, represents the total output of the kth region, n represents the number of regions included in the research scope; ΔM k A positive or negative value indicates that the consumption value of the region is more or less than the production value.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the static method for regional classification of embodied carbon emissions in a multi-regional environment as described in any one of claims 1 to 7.
