CN-122022853-A - Carbon footprint accounting method and system based on supply chain data tracing

CN122022853ACN 122022853 ACN122022853 ACN 122022853ACN-122022853-A

Abstract

The invention discloses a carbon footprint accounting method and system based on supply chain data tracing, and aims to solve the problem that the existing carbon footprint accounting method in the photovoltaic industry is low in accuracy; the method comprises the steps of drawing a photovoltaic product unit process model and collecting unit process activity level data according to basic information, process and supply chain information of a photovoltaic product, constructing a SENet-Conformer deep semantic network model based on an attention mechanism, splicing supply chain live-action data and matching carbon footprint factors based on the model to obtain supply chain splicing result data and carbon footprint factor data, generating a full life cycle carbon footprint accounting model, carrying out model verification on the accounting model, initially calculating a carbon footprint of the photovoltaic product, carrying out multidimensional anomaly detection on the result, evaluating data quality and analyzing uncertainty, and outputting a final carbon footprint result.

Inventors

HUANG WEI
Wu Mengxiaojun
LIN LANG
WEI DANQING
WANG CHENG
YUAN NINGJING
LU YIBING
ZHANG HONGYING
WANG YILUN

Assignees

浙江省经济信息中心(浙江省价格研究所)

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. The carbon footprint accounting method based on supply chain data tracing is characterized by comprising the following steps of: S1, drawing a photovoltaic product unit process model and collecting unit process activity level data according to basic information, process and supply chain information of a photovoltaic product; S2, performing supply chain live-action data splicing and carbon footprint factor matching on the basis of SENet-Conformer depth semantic network model of an attention mechanism to obtain supply chain splicing result data and carbon footprint factor data, and generating a full life cycle carbon footprint accounting model; And S3, performing model verification on the accounting model, primarily calculating the carbon footprint of the photovoltaic product, performing multidimensional abnormal verification on the result, evaluating the data quality and analyzing the uncertainty, and outputting a final carbon footprint result.
2. The supply chain data trace back-based carbon footprint accounting method as claimed in claim 1, wherein said step S1 comprises the steps of: s11, determining basic information of a photovoltaic product; s12, defining functional units or declaration units of the photovoltaic product, life cycle system boundaries, data collection time periods and trade-off criteria; s13, carding emission sources according to the range 1, the range 2 and the range 3 and identifying greenhouse gas types; s14, describing process information and upstream and downstream supply chain information of the photovoltaic product according to the production process of the photovoltaic product.
3. The supply chain data trace-back based carbon footprint accounting method according to claim 1 or 2, wherein the step S2 comprises the steps of: and determining multi-stage supply chain data of the photovoltaic product according to the step S1, wherein the photovoltaic product is composed of a plurality of first-stage supply chain data, the first-stage supply chain data comprise a plurality of first-stage intermediate photovoltaic raw material products, the first-stage intermediate photovoltaic raw material products are composed of a plurality of second-stage supply chain data, and the second-stage supply chain data comprise a plurality of second-stage intermediate photovoltaic raw material products until the supply chain data cannot be traced.
4. A supply chain data trace back-based carbon footprint accounting method according to claim 3, wherein said step S2 comprises the steps of: splicing the supply chain live-action data by using SENet-Conformer deep semantic network model; Judging whether the last-stage supply chain data of the photovoltaic product is searched, if so, splicing the supply chain live-action data, and continuing to trace the supply chain data upwards by the spliced intermediate photovoltaic raw material product until the supply chain data is traced back to the elementary stream; if the carbon footprint factor data cannot be matched, the carbon footprint factor data is manually input.
5. The supply chain data trace back-based carbon footprint accounting method as claimed in claim 1 or 2, wherein the SENet-Conformer deep semantic network model is improved on the basis of a DSSM frame, a attention introducing mechanism is adopted, a SENet module is added between a feature input layer and a mixed representation layer of the DSSM frame, a Conformer mixed architecture is adopted to replace an original DNN network in the mixed representation layer of the DSSM frame, and the Conformer mixed architecture comprises a CNN network and a Transformer.
6. A supply chain data trace back-based carbon footprint accounting method as claimed in claim 3, wherein said preliminary calculation of a photovoltaic product carbon footprint comprises: The total amount of greenhouse gas emission based on the functional unit is equal to the sum of greenhouse gas emission generated by an emission source directly controlled by an enterprise, indirect greenhouse gas emission generated by electric power and heating power purchased by the enterprise and greenhouse gas emission generated by upstream and downstream activities in a supply chain of the enterprise.
7. The supply chain data trace back-based carbon footprint accounting method as claimed in claim 1 or 6, wherein said multi-dimensional anomaly detection of results comprises: And performing Z-score probability statistical model fitting on the carbon footprint data of the corresponding category to obtain a carbon footprint statistical probability model of the category, performing Z-score calculation according to the carbon footprint result of the target photovoltaic product and the obtained probability model, and further judging the abnormal condition of the data.
8. The supply chain data trace back-based carbon footprint accounting method according to claim 7, wherein the abnormal condition of the data is specifically judged according to a Z-score absolute value obtained by calculation of the Z-score, if the Z-score absolute value is smaller than or equal to a set threshold value, the data is normal, quality evaluation and uncertainty analysis are performed, and if the Z-score absolute value is larger than the set threshold value, the data deviates significantly from an average value and is marked as abnormal data.
9. The supply chain data trace back-based carbon footprint accounting method according to claim 1 or 2, wherein the model verification comprises data integrity verification, model quality inspection and quality balance inspection.
10. A supply chain data trace-based carbon footprint accounting system adapted to the supply chain data trace-based carbon footprint accounting method of any one of claims 1-9, comprising: The information generation module is used for determining basic information of the photovoltaic product to be calculated; The model building module is used for carrying out supply chain live-action data splicing and carbon footprint factor data matching based on SENet-Conformer deep semantic network models to generate a full life cycle carbon footprint accounting model; The model verification module is used for verifying the full life cycle carbon footprint accounting model; and the calculation analysis module is used for performing preliminary calculation on the carbon footprint of the photovoltaic product, and performing multidimensional abnormal inspection, data quality evaluation and uncertainty analysis on the result.

Description

Carbon footprint accounting method and system based on supply chain data tracing Technical Field The invention relates to the technical field of carbon footprint accounting, in particular to a carbon footprint accounting method and system based on supply chain data tracing. Background Currently, the clean energy industry represented by the photovoltaic industry is rapidly growing, and the clean energy industry is a new engine for meeting the world energy demands and promoting the economic growth. Photovoltaic is taken as a new three-sample of foreign trade in China, and is subjected to various carbon barriers or invisible carbon barriers which are sequentially promoted by various countries in recent years, so that enterprises in the photovoltaic industry face serious challenges. Currently, due to the lack of a localized carbon footprint accounting tool and a background database which meet the international mutual recognition requirement of carbon footprint, the international carbon footprint background database only has a small amount of old and native carbon footprint factor data, and the adoption of the international carbon footprint background database can cause overestimation of the carbon footprint data of the photovoltaic products in China, so that the international market competitiveness of the domestic photovoltaic products is weakened. Therefore, in order to improve the accuracy of the carbon footprint result, the cradle-to-gate live-action data should be used for each supply chain raw and auxiliary material data. However, in actual accounting, the utilization rate of the real data of the supply chain is not high, most primary suppliers have difficulty in providing raw material cradle-to-gate data, only gate-to-gate enterprise data can be provided, and effective verification on an accounting result is lacking, in this case, a supply chain carbon footprint data splicing method based on a splicing idea is provided, so that the quality of the carbon footprint data and the accuracy of the accounting result are improved. According to the method, the supply chain data splicing method is preferentially used for splicing the data from the gate to the gate of the supplier until the supply chain data cannot be found or the supply chain data cannot be traced back to the elementary stream, and then the endmost raw and auxiliary material products are subjected to accounting by using the factor accounting method. The Chinese patent with the publication number of CN117522179A discloses a distributed energy life cycle carbon footprint research method which comprises the steps of dividing the full life cycle carbon footprint boundary range of distributed energy based on a life cycle evaluation theory, constructing a carbon footprint accounting model according to the divided life cycle carbon footprint boundary, carrying out life cycle carbon footprint accounting on distributed energy, and calculating to obtain the carbon emission recovery period of a distributed energy system in a park according to the carbon emission reduction amount and the carbon emission amount in the life cycle of the distributed energy system. Disclosure of Invention The invention solves the problems of low accuracy, low utilization rate of supply chain live-action data and lack of verification results of the existing carbon footprint verification method, provides a supply chain data traceable carbon footprint verification method and system, performs supply chain live-action data splicing and carbon footprint factor matching based on SENet-Conformer depth semantic network model of an attention mechanism to obtain supply chain splicing result data and carbon footprint factor data, generates a full-life-cycle carbon footprint verification model, remarkably improves accuracy of supply chain live-action data splicing and carbon footprint factor matching, and improves quality of carbon footprint data and accuracy of results. In order to achieve the purpose, the invention adopts the following technical scheme that the carbon footprint accounting method based on supply chain data tracing comprises the following steps: S1, drawing a photovoltaic product unit process model and collecting unit process activity level data according to basic information, process and supply chain information of a photovoltaic product; S2, performing supply chain live-action data splicing and carbon footprint factor matching on the basis of SENet-Conformer depth semantic network model of an attention mechanism to obtain supply chain splicing result data and carbon footprint factor data, and generating a full life cycle carbon footprint accounting model; And S3, performing model verification on the accounting model, primarily calculating the carbon footprint of the photovoltaic product, performing multidimensional abnormal verification on the result, evaluating the data quality and analyzing the uncertainty, and outputting a final carbon footprint result. According to the t