CN-121998481-A - Multi-level dynamic enterprise product carbon emission calculation method and device

CN121998481ACN 121998481 ACN121998481 ACN 121998481ACN-121998481-A

Abstract

The invention provides a multi-level dynamic enterprise product carbon emission calculation method and device, and relates to the technical field of data processing. The method comprises the steps of obtaining carbon emission related data of an enterprise, preprocessing the carbon emission related data of the enterprise to obtain a preprocessed data set, obtaining dynamic carbon emission factors by utilizing a multi-level carbon emission factor dynamic optimization model based on the preprocessed data set, calculating total carbon emission of the enterprise based on the dynamic carbon emission factors, determining product carbon emission factors by utilizing a product carbon emission regression model based on the total carbon emission of the enterprise, and representing unit carbon emission of products of the enterprise by the product carbon emission factors. The method and the system can flexibly adapt to different enterprises, fully utilize available data, reflect dynamic change of regional energy structures, and formulate an optimal carbon bank calculation scheme of the enterprises based on the dynamic change, thereby ensuring economy and scientificity of the carbon bank scheme of the enterprises, effectively reducing data calculation errors and improving calculation accuracy.

Inventors

Zhang Hongruo
SUN CHONG
FENG BO
LI MENGYU
LI JI
AN YAGANG

Assignees

国网河北省电力有限公司营销服务中心
国家电网有限公司

Dates

Publication Date: 20260508
Application Date: 20251229

Claims (10)

1. A multi-level dynamic enterprise product carbon emission calculation method, comprising: acquiring carbon emission related data of enterprises; preprocessing the related data of the carbon emission of the enterprise to obtain a preprocessed data set; based on the preprocessed data set, a dynamic carbon emission factor is obtained by utilizing a multi-level carbon emission factor dynamic optimization model; calculating a total amount of enterprise carbon emissions based on the dynamic carbon emission factor; And determining a product carbon emission factor by utilizing a product carbon emission regression model based on the total amount of carbon emission of the enterprise, wherein the product carbon emission factor characterizes the unit carbon emission of the product of the enterprise.
2. The method for calculating carbon emissions of a multi-level dynamic enterprise product of claim 1, wherein the obtaining the dynamic carbon emissions factor using a multi-level carbon emissions factor dynamic optimization model based on the preprocessed data set comprises: calculating an provincial fire power average carbon emission factor according to the preprocessed data set; Taking the provincial fire power average carbon emission factor as a reference, and combining electricity utilization and green electricity data of each region in the preprocessed data set to construct a multi-level carbon emission factor dynamic optimization model of district-county-station areas; and calculating the dynamic carbon emission factors of each city, county and district at different moments according to the multi-level carbon emission factor dynamic optimization model.
3. The method of claim 2, wherein said calculating a provincial fire power average carbon emission factor from said preprocessed data set comprises: according to the preprocessed data set, the provincial fire power average carbon emission factor is calculated by combining a first formula; the first formula includes: Wherein, the An average carbon dioxide emission factor for p power savings; An average carbon dioxide emission factor for p-fire power; The direct emission of carbon dioxide generated by power generation is saved for p; N-power-saving average carbon dioxide emission factors for net delivery of power to p-power-saving; the electric quantity which is sent out from the power saving to the power saving p is n; carbon dioxide emission factors are averaged for k national levels that save a net export power to p; The electric quantity exported to p is saved for k countries; an average carbon dioxide emission factor for regional power grid r; the electric quantity which is sent out from the regional power grid r to p is saved; The total electricity consumption of the year is p province; The method comprises the steps of generating energy by total new energy of p provinces, p is a target province, n is other provinces for sending electricity to p provinces, k is a country for exporting electricity to p provinces, and r is an regional power grid where p provinces are located.
4. The multi-level dynamic enterprise product carbon emission calculation method of claim 1, wherein the calculating the enterprise carbon emission total based on the dynamic carbon emission factor comprises: matching the enterprise to a corresponding dynamic carbon emission factor level according to the power supply attribute of the enterprise; Calculating indirect carbon emission generated by power consumption of the enterprise according to the power consumption data of the enterprise and a corresponding dynamic carbon emission factor level, wherein the dynamic carbon emission factor level comprises counties and areas, and the dynamic carbon emission factor of the area to which the enterprise belongs is calculated for the enterprise accessing the public transformer area; Acquiring energy consumption data of an enterprise; Calculating the direct carbon emission of the enterprise based on the energy consumption data of the enterprise; and determining the total carbon emission of the enterprise by integrating the indirect carbon emission and the direct carbon emission.
5. The method of claim 1, wherein determining the product carbon emission factor using a product carbon emission regression model based on the total amount of enterprise carbon emissions comprises: Taking output data of various products of an enterprise as independent variables and taking the total carbon emission of the enterprise as dependent variables, and constructing a product carbon emission regression model; And introducing physical range constraint of a product carbon emission coefficient into the product carbon emission regression model, taking the minimization of errors between the predicted total enterprise carbon emission and the actual total enterprise carbon emission as an optimization target, solving the product carbon emission regression model by adopting a gradient-based optimization algorithm, and outputting the product carbon emission factor.
6. The multi-level dynamic enterprise product carbon emission calculation method of claim 5, wherein the gradient-based optimization algorithm is a lagrangian multiplier method or a sequential quadratic programming algorithm.
7. The method for calculating carbon emission of a multi-level dynamic enterprise product according to claim 5, wherein the constructing a product carbon emission regression model using the output data of each type of product of the enterprise as an independent variable and the total amount of carbon emission of the enterprise as an independent variable comprises: taking output data of various products of an enterprise as independent variables, taking the total carbon emission of the enterprise as dependent variables, and constructing a product carbon emission regression model by combining a second formula; the second formula includes: Wherein Y is the total amount of carbon emission of the enterprise; The product yield of enterprises; is the unit carbon emission coefficient of the p-th type product, namely the carbon emission corresponding to the yield of each unit of the p-th type product; A collection of other factors that affect carbon emissions; is the coefficient of the O-th other influencing factors and represents the influence degree on the carbon emission; Is a random error term.
8. The method for computing carbon emissions for a multi-level dynamic enterprise product of claim 1, wherein the preprocessing the data related to carbon emissions for the enterprise to obtain a preprocessed data set comprises: identifying abnormal values in the carbon emission related data of the enterprise based on OneClassSVM algorithm, removing the identified abnormal values, and marking the abnormal values as blank values; and filling the blank value based on a K nearest neighbor classification algorithm to form the preprocessed data set.
9. The multi-level dynamic enterprise product carbon emission calculation method of claim 8, wherein the identifying outliers in the enterprise carbon emission related data based on OneClassSVM algorithm comprises: Constructing an electric quantity numerical data matrix according to the related data of the carbon emission of the enterprise; establishing a hypersphere model by utilizing OneClassSVM algorithm according to the electric quantity numerical data matrix; constructing a nonlinear objective function on the basis of the hypersphere model; optimizing the hypersphere model based on the nonlinear objective function to obtain an optimized hypersphere model; Calculating the projection distance from each data point of the electric quantity numerical data matrix to the center of the hypersphere model after optimization; and identifying an outlier based on the distance and the set outlier determination threshold.
10. A multi-level dynamic enterprise product carbon emission computing device, comprising: the data acquisition module is used for acquiring the carbon emission related data of the enterprise; the data preprocessing module is used for preprocessing the carbon emission related data of the enterprise to obtain a preprocessed data set; The dynamic carbon emission factor calculation module is used for obtaining dynamic carbon emission factors by utilizing a multi-level carbon emission factor dynamic optimization model based on the preprocessed data set; A total carbon emission calculation module for calculating a total carbon emission of the enterprise based on the dynamic carbon emission factor; And the product carbon emission factor calculation module is used for determining a product carbon emission factor by utilizing a product carbon emission regression model based on the total carbon emission amount of the enterprise, wherein the product carbon emission factor represents the unit carbon emission amount of the product of the enterprise.

Description

Multi-level dynamic enterprise product carbon emission calculation method and device Technical Field The invention relates to the technical field of data processing, in particular to a method, a device and equipment for calculating carbon emission of a multi-level dynamic enterprise product. Background The national carbon emission monitoring and analyzing service platform has the capability of checking national and regional carbon emission data in a month, and the overall technology reaches the international leading level. However, research and application of the current service in two key dimensions of regional dynamic carbon emission factors and enterprise-level carbon emission remain blank. The regional dynamic carbon emission factor is calculated, the indirect carbon emission of the consumption end is accurately calculated, the time sequence change of the carbon emission intensity of the power grid in different regions is revealed, and therefore a core data support is provided for local establishment of a scientific energy transformation path. And the enterprise-level carbon emission is used for quantifying, monitoring and analyzing the carbon emission condition of the enterprise so as to help the enterprise to formulate an emission reduction strategy and monitor the implementation condition. At present, research on carbon emission metering methods for enterprise products is particularly lacking. For example, the application number is 202211118933.1, and the patent name is 'energy consumption monitoring optimization method based on enterprise carbon measurement', and various energy data are sampled and analyzed through a special data acquisition module. Although the accuracy of the judgment of the monitoring period is improved, the method is highly dependent on customized hardware, is complex to implement and high in cost, and is difficult to popularize in enterprises of different scales and types. The related art mainly includes three ways of calculating carbon emission, namely an actual measurement method, a material balance algorithm and an emission coefficient method. The actual measurement method has high cost, large technical difficulty and limited representativeness, the material weighing algorithm depends on comprehensive and accurate material data, the acquisition is difficult, the calculation error is easy to accumulate and amplify, the emission coefficient method generally adopts static and universal emission coefficients, the dynamic change and the actual condition of a specific area or an enterprise production flow cannot be truly reflected, and systematic deviation exists in the calculation result. Therefore, there is a need to develop a product-level carbon emission calculation method that can flexibly adapt to different enterprises, make full use of available data, and reflect dynamic changes in regional energy structures. Disclosure of Invention The embodiment of the invention provides a multi-level dynamic enterprise product carbon emission calculation method, device and equipment, which can be flexibly adapted to different enterprises, fully utilize available data and reflect regional energy structure dynamic changes. In a first aspect, an embodiment of the present invention provides a method for calculating carbon emissions of a multi-level dynamic enterprise product, including: acquiring carbon emission related data of enterprises; Preprocessing the related data of the carbon emission of the enterprise to obtain a preprocessed data set; based on the preprocessed data set, a dynamic carbon emission factor is obtained by utilizing a multi-level carbon emission factor dynamic optimization model; calculating total carbon emission of the enterprise based on the dynamic carbon emission factor; And determining a product carbon emission factor by utilizing a product carbon emission regression model based on the total amount of the enterprise carbon emission, wherein the product carbon emission factor represents the unit carbon emission of the product of the enterprise. In a possible implementation manner of the first aspect, the obtaining the dynamic carbon emission factor using the multi-level carbon emission factor dynamic optimization model based on the preprocessed data set includes: Calculating the average carbon emission factor of the provincial fire according to the preprocessed data set; taking the provincial fire power average carbon emission factor as a reference, and combining the electricity utilization and green electricity data of each region in the preprocessed data set to construct a dynamic optimization model of the regional city-county-district multi-level carbon emission factor; And calculating the dynamic carbon emission factors of each city, county and district at different moments according to the multi-level carbon emission factor dynamic optimization model. In a possible implementation manner of the first aspect, the provincial fire power average carbon emission fact