Macroscopic carbon emission monitoring model based on'electricity-carbon calculation 'model
Through the macro carbon emission monitoring model based on the "electric-carbon calculation" model, the problems of inaccurate carbon emission accounting methods and limitations of monitoring technology in my country have been solved, and the rapid and accurate accounting and real-time monitoring of carbon emissions have been achieved, providing strong technical support for the realization of the "dual carbon" goal.
Patent Information
- Application Number
- CN202411797013.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-09
- Publication Date
- 2025-04-25
AI Technical Summary
In the process of promoting the "dual carbon" work, my country faces problems such as inaccurate carbon emission accounting methods and limitations in monitoring technology, and it is difficult to achieve rapid and accurate accounting and real-time monitoring of carbon emissions.
A macro carbon emission monitoring model based on the "electric-carbon computing" model is adopted. This model calculates energy consumption and industrial output data through regression analysis model, and multiplies it by carbon emission factor to calculate carbon emissions. Combined with the autoregressive distribution lag model (ARDL) and the minimum error optimization algorithm, it realizes rapid and accurate calculation of carbon emissions.
It has achieved rapid and accurate accounting and real-time monitoring of carbon emissions, improved the accuracy and reliability of carbon emission data, and provided technical support for the realization of the "dual carbon" goal.
Smart Images
Figure CN119886523A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of carbon emission monitoring, and in particular to a macroscopic carbon emission monitoring model based on an "electricity-carbon calculation" model. Background Art
[0003] First, my country's carbon emissions have long been overestimated by foreign research institutions, which has had an adverse impact on my country's international negotiations and international trade. Therefore, quickly and accurately calculating carbon emissions has become an important basis for carrying out "dual carbon" work. However, the current carbon emission accounting methods have many shortcomings, such as limited coverage, weak timeliness, and insufficient resolution. These problems seriously restrict the accuracy and reliability of carbon emission data.
[0004] Secondly, in the field of carbon emission monitoring technology, although countries have increased their investment in promoting technological innovation in recent years, the existing monitoring technology still has certain limitations. Traditional monitoring methods often find it difficult to achieve high-frequency and high-spatial-resolution carbon emission monitoring in small areas such as industrial parks and cities, and emerging technologies such as drone-coordinated carbon observation networks, online monitoring systems, and satellite monitoring have improved monitoring accuracy and coverage to a certain extent, but there are still many technical bottlenecks and practical application problems.
[0005] In response to the above problems, the industry has begun to explore new carbon emission accounting and monitoring methods. Among them, the "electricity-carbon calculation model" that "calculates energy with electricity and carbon with energy" through energy big data has become an effective solution. This model uses big data technology and algorithms to conduct in-depth mining and analysis of energy data, realizes accurate accounting and real-time monitoring of carbon emissions, and effectively solves the problems existing in traditional accounting methods. However, the model still needs to be further improved and optimized in practical applications to improve its accuracy and reliability.
[0006] To sum up, in the process of promoting the "dual carbon" work, my country faces problems such as inaccurate carbon emission accounting methods and insufficient monitoring technology. It is urgent to explore new solutions and technical means to achieve rapid and accurate accounting and real-time monitoring of carbon emissions. Summary of the invention
[0007] The problem to be solved by the present invention is: in the process of promoting the "dual carbon" work, in response to the problems of inaccurate carbon emission accounting methods and limitations of monitoring technology in my country, a new solution and technical means are explored and provided to achieve rapid and accurate accounting and real-time monitoring of carbon emissions, thereby providing strong technical support for the realization of my country's "dual carbon" goals.
[0008] The technical solution adopted by the present invention is: a macro carbon emission monitoring model based on the "electricity-carbon calculation" model, characterized in that the "electricity-carbon calculation" model trains historical electricity, GDP, clean energy share data with energy consumption and industrial output data, and calculates carbon emissions by constructing a regression analysis model, including two major steps:
[0009] (1) Calculating energy with electricity: The monthly electricity consumption data is brought into the constructed regression analysis model, and the monthly energy consumption and industrial output data are calculated through mathematical operations;
[0010] The regression analysis model is trained based on the correlation between historical electricity, GDP, clean energy share data and energy consumption and industrial output data to ensure the accuracy and reliability of the calculation;
[0011] (2) Calculate carbon based on energy: After obtaining energy consumption and industrial output data, multiply them by the corresponding carbon emission factor to obtain the carbon emissions for that month;
[0012] Carbon emission factors include general factors and dynamic electricity carbon emission factors, which are determined and adjusted according to specific energy types, industrial production processes, and dynamic changes in electricity data;
[0013] An autoregressive distributed lag model ARDL of power industry data and energy consumption data was constructed, and the current energy data was quickly calculated through the current electricity data. Among them, the dependent variable is the historical energy consumption data, the explanatory variables are the historical and current electricity data, and the supplementary variables are the proportion of clean energy and GDP influencing factors. The model takes electricity as the main input, and at the same time fully considers the impact of the development of my country's energy structure, consumption structure, and economic structure on carbon emissions, and comprehensively calculates the energy consumption data.
[0014] As a further solution of the present invention: the decomposition of influencing factors consists of three parts: determining the components of influencing factors, configuring the weights of influencing factors, and solving the optimization problem; using the minimum error optimization algorithm, by quantifying the variables and their weights that affect the energy consumption of cities and states, the historical energy consumption data of cities and states are estimated.
[0015] As a further solution of the present invention: the components of the influencing factors are determined: according to the input variables of the macro "electricity-carbon calculation model", electricity consumption, clean energy power generation data, and tertiary industry GDP economic data are used as influencing factors of energy consumption; based on the energy structure and industrial structure characteristics of each city and state, combined with the energy data and population data published in the statistical yearbook, the population size of the prefecture-level city, the energy consumption index per unit GDP, and the growth rate of energy consumption per unit GDP are added as influencing factors of the model.
[0016] As a further solution of the present invention: the influencing factor weight configuration: adopt a linear regression algorithm to fit the initial parameters of each influencing factor and energy consumption, multiply the initial parameters by the unit influencing factor energy consumption conversion coefficient to obtain the initial weight of each influencing factor, normalize the initial weight, and finally determine the weight of each influencing factor.
[0017] As a further solution of the present invention: Solving the optimization problem: The minimum error optimization algorithm is a typical quadratic programming equality constraint problem. A city and state energy consumption solution function is constructed, and the Lagrangian method is used to solve the optimization problem to obtain the annual energy consumption data of each city.
[0018] As a further solution of the present invention: the electric energy calculation is the main part of the model, which includes two sub-modules; module 1 is a regression analysis calculation, which uses electricity data regression analysis to calculate energy activities and industrial production process output data; module 2 is a monthly split calculation, which splits the annual low-frequency data into monthly high-frequency data.
[0019] As a further solution of the present invention: the carbon calculated by energy is "calculating carbon by energy" which multiplies the energy consumption and industrial output data output by "calculating energy by electricity" by the corresponding carbon emission factor to obtain carbon emissions; the carbon emission factors include the general factors defined in the "Guidelines for the Compilation of Provincial Greenhouse Gas Inventories" and the "Guidelines for Enterprise Greenhouse Gas Emissions Accounting Methods and Reporting", as well as the dynamic electricity carbon emission factors calculated by the macro "electricity-carbon calculation model".
[0020] Beneficial effects of the present invention:
[0021] The prefecture-level "electricity-carbon calculation model" has been deployed and applied in the Sichuan Province Carbon Emission Monitoring and Analysis Service Platform, which can realize regional carbon emission monitoring of 21 cities (prefectures), carbon emission monitoring of 7 major industries and 7 sub-industries of industry, and provide data support for the government to clearly grasp the macro situation of regional carbon emissions and policy formulation. At present, 21 cities (prefectures) have completed the deployment and pilot application of the prefecture-level "electricity-carbon calculation model", and verified the calculation results of the first batch of 6 pilot cities. The calculation results of the prefecture-level cities are consistent with the provincial calculation results, indicating that the prefecture-level "electricity-carbon calculation model" has strong stability, reliability and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 This is a flow chart of a prefecture-level "electricity-carbon calculation model" of a macro carbon emission monitoring model based on the "electricity-carbon calculation" model of the present invention.
[0023] Figure 2 The invention discloses a minimum error optimization algorithm model for a macro carbon emission monitoring model based on an "electricity-carbon calculation" model.
[0024] Figure 3The present invention discloses a Lagrangian function model of a macro carbon emission monitoring model based on an "electricity-carbon calculation" model.
[0025] Figure 4 The present invention discloses an electric calculation energy flow chart of a macro carbon emission monitoring model based on the "electricity-carbon calculation" model.
[0026] Figure 5 The present invention discloses a flow chart of calculating carbon based on energy for a macro carbon emission monitoring model based on the "electricity-carbon calculation" model.
[0027] Figure 6 This is a comparative analysis chart of Sichuan Province's carbon emission data from 2007 to 2021 based on a macro carbon emission monitoring model based on the "electricity-carbon calculation" model of the present invention.
[0028] Figure 7 This is a model scheduling flow chart of a macro carbon emission monitoring model based on the "electricity-carbon calculation" model of the present invention. DETAILED DESCRIPTION
[0029] The present invention will be further described below.
[0030] Feasibility analysis: From the perspective of physical mechanism, energy activities account for 88% of the total carbon emissions of the whole society, of which 41% comes from electricity, so electricity carbon emissions account for 36% of the total carbon emissions. The power industry has a high proportion of carbon emissions. At the same time, electricity is the blood and lifeline of social and economic development, covering all aspects of residents' lives, industrial production, etc., and has a physical correlation with economic and social activities and carbon emissions of the whole society. Electricity carbon calculation is feasible in terms of physical mechanism. From the perspective of mathematical mechanism, electricity consumption has a significant strong correlation with energy activities and carbon emissions in industrial production processes (0.955, 0.810), and also has a significant strong correlation with total carbon emissions (0.951). Therefore, energy activities and industrial production process data can be calculated through electricity data, so as to quickly and accurately calculate carbon emissions. Electricity carbon calculation is feasible in terms of mathematical mechanism. From the perspective of data advantages, electricity data has achieved comprehensive automated collection, transmission and processing, and has the advantages of strong real-time, high accuracy, high resolution and wide collection range. It can provide a high-quality basic data source for carbon emission accounting, and electricity carbon calculation is feasible in terms of data.
[0031] A macro carbon emission monitoring model based on the "electricity-carbon calculation" model. The "electricity-carbon calculation" model trains historical electricity, GDP, clean energy share and other data with energy consumption and industrial output data, and calculates carbon emissions by building a regression analysis model. It includes two steps:
[0032] (1) Calculating energy with electricity: The monthly electricity consumption data is brought into the constructed regression analysis model, and the energy consumption and industrial output data of the month are calculated through mathematical operations.
[0033] The regression analysis model is trained based on the correlation between historical electricity consumption, GDP, clean energy share and other data and energy consumption and industrial output data to ensure the accuracy and reliability of the calculation.
[0034] (2) Calculate carbon based on energy: After obtaining the energy consumption and industrial output data, multiply them by the corresponding carbon emission factor to obtain the carbon emissions for that month.
[0035] Carbon emission factors include general factors and dynamic electricity carbon emission factors, which are determined and adjusted based on specific energy types, industrial production processes, and dynamic changes in electricity data.
[0036] An autoregressive distributed lag model (ARDL) of power industry data and energy consumption data was constructed, and the current energy data was quickly calculated through the current electricity data, solving the problem of insufficient timeliness of energy data statistics. Among them, the dependent variable is the historical energy consumption data, the explanatory variables are the historical and current electricity data, and the supplementary variables are the proportion of clean energy, GDP and other influencing factors. The model uses electricity as the main input, and fully considers the impact of my country's energy structure, consumption structure, and economic structure development on carbon emissions, and comprehensively calculates energy consumption data. Figure 1 shown.
[0037] In response to the difficult problems of uneven statistical quality and inconsistent statistical caliber among various cities (states), we conducted in-depth research on the data on electricity, economy, population, energy consumption, official carbon emissions, etc. of each city (state), and clarified the data sources (online power grid, provincial statistical yearbooks, city (state) statistical yearbooks, etc.) and update frequency.
[0038] Table 1 Multi-category data collection indicators and frequency table
[0039]
[0040]
[0041] The influencing factor decomposition consists of three parts: determining the influencing factor components, configuring the influencing factor weights, and solving the optimization problem. The minimum error optimization algorithm is used, such as Figure 2 As shown in Figure 1, by quantifying the variables and their weights that affect the city's (state's) energy consumption, the city's (state's) historical energy consumption data is estimated.
[0042] The components of the influencing factors are determined as follows: According to the input variables of the macro "electricity-carbon calculation model", electricity data such as electricity consumption and clean energy power generation, and economic data such as the GDP of the tertiary industry are used as influencing factors of energy consumption; based on the characteristics of the energy structure and industrial structure of each city (state), combined with the energy data and population data released in the statistical yearbook, the number of prefecture-level cities' population, energy consumption index per unit GDP, and energy consumption growth rate per unit GDP are added as influencing factors of the model.
[0043] Table 2 Factors affecting economic development and related variables
[0044]
[0045] The weight configuration of the influencing factors: adopt a linear regression algorithm to fit the initial parameters of each influencing factor and energy consumption, multiply the initial parameters by the unit influencing factor energy consumption conversion coefficient to obtain the initial weight of each influencing factor, normalize the initial weight, and finally determine the weight of each influencing factor.
[0046] (I) Calculation of initial parameters of influencing factors
[0047] The linear regression algorithm was used to calculate the initial parameters of each influencing factor in Sichuan Province.
[0048] 2. Determine the initial weight of each influencing factor
[0049] The energy consumption under the unit influencing factor is calculated as the conversion coefficient of each influencing factor, and the parameter value of each influencing factor is multiplied by the conversion coefficient to obtain the initial weight of each influencing factor.
[0050] Assuming that there are four influencing factors, the relationship between the GDP of the primary, secondary and tertiary industries, population and energy consumption is fitted through the linear regression algorithm, and the initial parameters of the GDP of the primary, secondary and tertiary industries, population and energy consumption are obtained, which are s 1 =0.4532, s 2 =0.2637, s 3 =0.5405,s 4 =0.4095, and the initial weights of the GDP of the primary, secondary and tertiary industries and the population are obtained by multiplying them with the conversion coefficient in step 2. After normalization, the final weights are Υ 1 =0.1328, Υ 2 =0.319, Υ 3 =0.3941, Υ 4 =0.1541.
[0051] (III) Normalization to determine the final weight
[0052] The initial weights of each influencing factor are normalized so that their sum is 1 to obtain the final weight of each influencing factor.
[0053] The optimization problem solution: The minimum error optimization algorithm is a typical quadratic programming equality constraint problem. The city (state) energy consumption solution function is constructed and the Lagrangian method is used to solve the optimization problem, such as Figure 3 As shown in Figure 2, we can obtain the annual energy consumption data of various cities.
[0054] The calculation of energy (output) by electricity is the main part of the model, which includes two submodules. Module 1 is regression analysis calculation, which uses data such as electricity for regression analysis to calculate energy activities and industrial production process output data; Module 2 is monthly split calculation, which splits annual low-frequency data into monthly high-frequency data, such as Figure 4 shown.
[0055] The carbon calculation based on energy is to calculate carbon by multiplying the energy consumption and industrial output data output by "calculating energy (output) based on electricity" by the corresponding carbon emission factor to obtain carbon emissions. Carbon emission factors include the general factors defined in the "Guidelines for the Preparation of Provincial Greenhouse Gas Inventories" and the "Guidelines for the Calculation and Reporting of Greenhouse Gas Emissions for Enterprises (Trial)", as well as the dynamic electricity carbon emission factors calculated by the macro "electricity-carbon calculation model", such as Figure 5 shown.
[0056] Model verification. At the regional level, compared with the provincial authoritative carbon emission data source from 2005 to 2021, the average absolute error is 1.07%, the maximum error is -2.44%, and the minimum error is 0.91%. The calculation result verification of the prefecture-level "electricity-carbon calculation model" is divided into overall comparison verification and partial comparison verification. The overall comparison verification is to compare and analyze the model calculation results with the Sichuan carbon emission data provided by the Sichuan Provincial Development and Reform Commission. The sum of the carbon emissions from energy activities of the 21 cities (prefectures) calculated by the model from 2007 to 2021 is compared with the Sichuan carbon emission data from 2007 to 21 provided by the Provincial Development and Reform Commission (such as Figure 6 In the past 15 years, the lowest deviation rate was 0.09%, the highest deviation rate was 2.78%, and the average deviation rate was only 1.07%, all of which were within the 95% confidence interval.
[0057] Part of the comparative verification is to compare and analyze the model calculation results with the carbon emission data of cities and prefectures provided by the municipal (state) government. The comparative data include the carbon emissions of Yibin City calculated by the model in 2018 and 2020, and the carbon emission data provided by the Yibin Municipal Ecological Environment Bureau. The lowest deviation rate of the six data in the two years is 0.02%, the highest deviation rate is 1.40%, and the average deviation rate is only 0.83%. From the results of the comparative verification, it can be seen that the model calculation data has a high accuracy.
[0058] Table 3 Carbon emission model calculation table for 2018 and 2020
[0059]
[0060]
[0061] The platform adopts an automatic model algorithm scheduling optimization strategy for the shortest job. For the containerized electricity-carbon computing models of 21 cities and states, full automation of computing processes and scheduling is achieved. The mechanism automatically performs model resource scheduling, calculation startup, result data acquisition, statistical collation and database updates on a monthly basis, effectively ensuring the efficient use of computing resources. The shortest job first (SJF) algorithm is used for dynamic priority scheduling to achieve synchronous calling of multiple models, improve model computing efficiency by 80%, increase throughput by 85%, shorten processing time by 50%, and improve resource scheduling efficiency by 75%. At the same time, model version management and rollback mechanisms are added to ensure the stability of model updates. Figure 7 shown.
[0062] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model, characterized by: The "electricity-carbon calculation" model trains historical electricity, GDP, clean energy share data with energy consumption and industrial output data, and calculates carbon emissions by building a regression analysis model, which includes two steps: (1) Calculating energy with electricity: The monthly electricity consumption data is brought into the constructed regression analysis model, and the monthly energy consumption and industrial output data are calculated through mathematical operations; The regression analysis model is trained based on the correlation between historical electricity, GDP, clean energy share data and energy consumption and industrial output data to ensure the accuracy and reliability of the calculation; (2) Calculate carbon based on energy: After obtaining energy consumption and industrial output data, multiply them by the corresponding carbon emission factor to obtain the carbon emissions for that month; Carbon emission factors include general factors and dynamic electricity carbon emission factors, which are determined and adjusted according to specific energy types, industrial production processes, and dynamic changes in electricity data; An autoregressive distributed lag model ARDL of power industry data and energy consumption data was constructed, and the current energy data was quickly calculated through the current electricity data. Among them, the dependent variable is the historical energy consumption data, the explanatory variables are the historical and current electricity data, and the supplementary variables are the proportion of clean energy and GDP influencing factors. The model takes electricity as the main input, and at the same time fully considers the impact of the development of my country's energy structure, consumption structure, and economic structure on carbon emissions, and comprehensively calculates the energy consumption data.
2. A macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model according to claim 1, characterized in that: The influencing factor decomposition consists of three parts: determining the influencing factor components, configuring the influencing factor weights, and solving the optimization problem. The minimum error optimization algorithm is used to quantify the variables and their weights that affect the city and state energy consumption, and estimate the city and state historical energy consumption data.
3. A macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model according to claim 2, characterized in that: The components of the influencing factors are determined as follows: according to the input variables of the macro "electricity-carbon calculation model", electricity consumption, clean energy power generation data, and tertiary industry GDP economic data are used as influencing factors of energy consumption; based on the characteristics of the energy structure and industrial structure of each city and state, combined with the energy data and population data released in the statistical yearbook, the number of prefecture-level cities' population, energy consumption index per unit GDP, and energy consumption growth rate per unit GDP are added as influencing factors of the model.
4. A macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model according to claim 3, characterized in that: The weight configuration of the influencing factors: adopt a linear regression algorithm to fit the initial parameters of each influencing factor and energy consumption, multiply the initial parameters by the unit influencing factor energy consumption conversion coefficient to obtain the initial weight of each influencing factor, normalize the initial weight, and finally determine the weight of each influencing factor.
5. A macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model according to claim 4, characterized in that: The optimization problem solution: The minimum error optimization algorithm is a typical quadratic programming equality constraint problem. It constructs a city and state energy consumption solution function, and uses the Lagrangian method to solve the optimization problem to obtain the annual energy consumption data of each city.
6. According to claim 1, a macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model is characterized in that: The electric energy calculation is the main part of the model, which includes two sub-modules: Module 1 is a regression analysis calculation, which uses electricity data regression analysis to calculate energy activities and industrial production process output data; Module 2 is a monthly split calculation, which splits the annual low-frequency data into monthly high-frequency data.
7. The macroscopic carbon emission monitoring model based on the "electricity-carbon calculation" model according to claim 1 is characterized in that: The so-called carbon calculation based on energy is to multiply the energy consumption and industrial output data output by "calculating energy based on electricity" by the corresponding carbon emission factor to obtain carbon emissions; the carbon emission factors include the general factors defined in the "Guidelines for the Preparation of Provincial Greenhouse Gas Inventories" and the "Guidelines for Enterprise Greenhouse Gas Emissions Accounting Methods and Reporting", as well as the dynamic electricity carbon emission factors calculated by the macro "electricity-carbon calculation model".