Supply chain carbon footprint tracing method for digital identity and verifiable certificate

Through the method of digital identity and verifiable credentials, combined with knowledge graph and blockchain technology, the problem of low data acquisition and collaboration efficiency in product carbon footprint accounting is solved, and the accurate prediction and full-process management of carbon footprint are achieved, which has improved the The carbon footprint traceability and data sharing transparency of the supply chain.

CN119067682APending Publication Date: 2024-12-03SHANGHAI CARBON BALANCE TECHNOLOGY CO LTD
View PDF 0 Cites 9 Cited by

Patent Information

Application Number
CN202411120949.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The existing carbon footprint accounting methods for product are facing problems such as difficulty in obtaining data, less knowledge accumulation, low coordination efficiency and weak traceability. Especially when the dynamics and uncertainties of the carbon footprint of product are high, it is difficult for traditional methods to achieve accuracy. Predictive and quantitative analysis of the entire life cycle.

Method used

Using the supply chain carbon footprint traceability method of digital identity and verifiable credentials, ontological representation learning is carried out through the knowledge graph, and combining blockchain technology to build a verifiable carbon footprint data evidence chain, and integrating knowledge graph and supervised learning paradigm to build LKGC -LSTM carbon footprint prediction model to realize quantitative prediction of product carbon footprint and full-process closed-loop management.

Benefits of technology

It improves the accuracy and interpretability of carbon footprint accounting, enhances the synergy efficiency and traceability of supply chain data, provides credible carbon footprint data sharing and openness and transparency, and supports green supply chain governance and low-carbon product promotion.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119067682A_ABST
    Figure CN119067682A_ABST
Patent Text Reader

Abstract

The supply chain carbon footprint tracing method based on the digital identity and the verifiable certificate comprises the following steps: step 1, constructing a product carbon footprint knowledge graph based on an ontology; 2, performing representation learning on the knowledge graph through a hierarchical relation attention network HRGAT to obtain a low-dimensional real value vector; 3, an LKGC-LSTM carbon footprint prediction model is constructed; step 4, the production enterprise registers an account and generates a unique digital identity, registers a product identity label and associates an enterprise identity; 5, the production enterprise submits a carbon footprint auditing application to the accounting mechanism; step 6, carrying out carbon footprint modeling and data acquisition; 7, verifying the authenticity of the voucher; 8, an accounting mechanism completes accounting and generates a carbon footprint report; step 9, the accounting mechanism completes the carbon footprint accounting certificate, and generates a verifiable carbon footprint certificate after using the digital identity signature; and step 10, the production enterprise obtains the carbon footprint voucher, and the carbon footprint voucher can be disclosed to a verifier or uploaded to a carbon disclosure platform for verification and traceability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain carbon footprint tracing methods, and more specifically, to a supply chain carbon footprint tracing method based on digital identity and verifiable credentials. Background Art

[0002] Product carbon footprint refers to the total amount of greenhouse gases emitted directly and indirectly during the entire life cycle of a product. It is an important tool for assessing the low-carbon attributes of products and promoting the development of green supply chains.

[0003] At present, there are many standards and initiatives at home and abroad involving product carbon footprint assessment, such as ISO / TS14067, PAS2050, GHG Protocol, etc. These standards provide guidance for product carbon footprint accounting in terms of accounting boundaries, data quality, emission factors, allocation methods, etc. However, in actual applications, there are still problems such as difficult data collection, low accounting efficiency, and poor supply chain coordination. This is mainly reflected in:

[0004] 1. It is difficult to obtain carbon footprint accounting data. Product carbon footprint accounting requires the collection of a large amount of activity level data such as raw material input, energy consumption, etc., and there are many types of data, the data sources are scattered, and the company's own data accumulation is insufficient, making it difficult to efficiently and accurately collect the data required for accounting;

[0005] 2. Insufficient accumulation and application of product carbon footprint knowledge. The carbon emission characteristics of different product types, process routes, and supply chain structures vary greatly. There is a lack of converting previous accounting knowledge and experience into a knowledge base with semantic associations and machine comprehension, resulting in a high reliance on the field experience of practitioners in the accounting process and a lack of unified and standardized accounting processes and rules;

[0006] 3. Lack of a coordinated mechanism for carbon footprint accounting across the entire supply chain. Product carbon footprint accounting requires upstream and downstream companies in the supply chain to provide original data from all links. However, due to inconsistent data standards and high communication costs, data circulation and coordination efficiency across companies is low. In addition, the carbon emission data provided by upstream suppliers lacks a credible verification mechanism and is difficult to be directly used for carbon footprint accounting by downstream companies.

[0007] 4. Product carbon footprint traceability and information disclosure need to be strengthened. Consumers, investors and other stakeholders urgently need to understand product carbon footprint information to support green consumption and investment decisions. However, the current product carbon footprint information disclosure channels are not smooth, lack of credible certification and traceability mechanisms, and the transparency of carbon footprint data needs to be improved.

[0008] Most existing methods are based on the traditional life cycle assessment (LCA) method to carry out carbon footprint accounting. They manually collect direct and indirect emission data of products, calculate carbon emissions based on emission factors, and compile carbon footprint reports. However, such methods are easily affected by factors such as data quality and accuracy of emission factors, and lack the ability to quantitatively predict and analyze emissions throughout the product life cycle. Some studies have proposed the use of emerging technologies such as blockchain for carbon emission data management, such as using blockchain distributed ledgers to solve problems such as carbon data rights confirmation, storage, and sharing. However, these methods mostly focus on the field of carbon emission rights trading, and do not consider the needs of cross-enterprise collaboration, knowledge utilization, and security traceability in product carbon footprint scenarios. For the quantitative assessment of product carbon footprint, most traditional methods are based on process LCA, which analyzes the carbon emission inventory of each link in the product life cycle by constructing a process tree based on input and output. However, in practice, the carbon footprint of a product is affected by many factors and has great dynamics and uncertainty. The static process LCA method is difficult to continuously predict and scenario analyze the carbon footprint of the product life cycle. In recent years, with the development of technologies such as the Internet of Things and big data, some scholars have tried to use machine learning methods to predict product carbon footprints. Some scholars have used support vector machines to establish product carbon footprint prediction models; others have used artificial neural networks to achieve quantitative assessment of carbon emissions during the product life cycle. However, these methods mainly build prediction models from the perspective of the correlation of emission data, lacking the integration of knowledge in areas such as product architecture, manufacturing processes, and environmental impact mechanisms, and the prediction accuracy and generalization performance need to be strengthened.

[0009] In summary, the existing product carbon footprint accounting and management methods still face problems such as difficulty in data acquisition, insufficient knowledge accumulation, low collaborative efficiency, and weak traceability. It is urgent to explore a new paradigm of knowledge-driven, data-enabled, and process-coordinated development around carbon footprint scenarios. Summary of the invention

[0010] The purpose of the present invention is to provide a supply chain carbon footprint traceability method with digital identity and verifiable credentials in order to solve the above problems. The present invention starts from cutting-edge technologies such as knowledge graph, blockchain, trusted digital identity, verifiable credentials, graph neural network, etc., to build a full-process closed-loop solution for product carbon footprint accounting, certification, and disclosure. On the one hand, the knowledge graph is introduced to ontologically and semantically represent the knowledge in the field of product carbon footprint, and the knowledge graph semantics is extracted using representation learning technology to guide the entire process of carbon footprint data collection and accounting; on the other hand, blockchain technology is used to build a verifiable, traceable, and tamper-proof carbon footprint data evidence chain to achieve trusted transmission and sharing of carbon footprint data of upstream and downstream companies; and the knowledge graph and supervised learning paradigm are innovatively integrated to build a quantitative prediction and analysis model for product carbon footprint.

[0011] The present invention provides a supply chain carbon footprint traceability method with digital identity and verifiable credentials, comprising the following steps: Step 1: construct a product carbon footprint knowledge graph based on the ontology to obtain entities, relationships and constraints at each product life cycle stage; Step 2: perform representation learning on the knowledge graph through a hierarchical relational attention network HRGAT to obtain a low-dimensional real-valued vector; Step 3: integrate the knowledge graph embedding vector into a long short-term memory network LSTM to construct an LKGC-LSTM carbon footprint prediction model; Step 4: the manufacturer registers an account and generates a unique digital identity, registers a product identity and associates it with the enterprise identity; Step 5: the manufacturer Submit a carbon footprint audit application to the accounting agency; Step 6: The accounting agency registers an account and generates a digital identity, accepts the audit application and conducts carbon footprint modeling and data collection; Step 7: The accounting agency collects verifiable carbon footprint certificates from supply chain companies and calls blockchain services to verify the authenticity of the certificates; Step 8: The accounting agency integrates the LKGC-LSTM carbon footprint prediction model to complete the accounting and generate a carbon footprint report; Step 9: The accounting agency completes the carbon footprint accounting certificate, signs it with a digital identity, and generates a verifiable carbon footprint certificate; Step 10: The production enterprise obtains the carbon footprint certificate, which can be disclosed to the verifier or uploaded to the carbon disclosure platform for verification and traceability.

[0012] Specifically, the steps of constructing a product carbon footprint knowledge graph based on ontology in step 1 include: step 1.1: based on the OWL ontology description language, define the core concepts, attributes and relationships of each link in the product life cycle, and build a TBox knowledge base; step 1.2: use natural language processing and data mining technology to extract entities and relationships from structured and unstructured data, and build an ABox assertion library; step 1.3: combine TBox and ABox to form an ALCQ (D) description logic knowledge base, and use JenaAPI to serialize it into an RDF format knowledge graph.

[0013] Specifically, the step of performing representation learning on the knowledge graph by using HRGAT in step 2 includes:

[0014] Step 2.1: Randomly initialize the entity embedding vector u according to the knowledge graph G = (E, R) i ∈R d and the relation embedding vector v r ∈R d ;

[0015] Step 2.2: For an entity type t∈T, compute the type embedding w by aggregating the entity embeddings belonging to t through attention t :

[0016]

[0017] Among them, σ is the activation function, f(·) is the attention scoring function; similarly, the embedding w of the relation type s∈S can be obtaineds ;

[0018] Step 2.3: Iteratively update the entity embedding h through the hierarchical attention mechanism i :

[0019]

[0020] in, is the attention weight of entity i and type t; is the attention weight of entity j on entity i via relation s;

[0021] Step 2.4: Repeat step 2.3 until the entity embedding converges and output the final embedding representation.

[0022] Specifically, the steps of constructing the LKGC-LSTM carbon footprint prediction model in step 3 include: Step 3.1: transforming the observed value x at the i-th time step into i Mapped to LSTM input vector input i =W i x i +b i ∈R k ; Step 3.2: Embed the knowledge graph into vector u i With input i Splice, get

[0023] Step 3.3: Based on Perform LSTM state transfer:

[0024]

[0025]

[0026] h i =o i tanh(C i )

[0027] Among them, f i 、i i , o i They are forget gate, input gate, and output gate respectively. is the candidate state, W and b are the parameter matrix and bias term;

[0028] Step 3.4: Use MLP to convert the LSTM final hidden state h t Mapped to the predicted carbon footprint value for the next N steps

[0029] Specifically, the step of registering an account and generating a unique digital identity by the manufacturing enterprise in step 4 includes:

[0030] Step 4.1: The enterprise submits registration information including enterprise name, unified social credit code, username and password;

[0031] Step 4.2: The system verifies the enterprise registration information and opens an account for the enterprise user after the application is approved;

[0032] Step 4.3: Generate enterprise digital identity based on public and private keys through the did:key mode, public key and digital identity;

[0033] Step 4.4: Use blockchain smart contracts to register the enterprise account address and digital identity information to the distributed identity registration center.

[0034] Specifically, the steps of registering the product identity and associating the enterprise identity in step 4 include:

[0035] Step 4.5: The enterprise submits basic product information such as name and model, and the system generates a digital product identity;

[0036] Step 4.6: Generate a unique ID for each product entity using the IoT electronic tag and bind it to its digital identity;

[0037] Step 4.7: Register the association between the product digital identity and the enterprise identity to the blockchain identity registration center.

[0038] Specifically, the step of collecting the verifiable carbon footprint certificates of supply chain enterprises by the accounting institution in step 7 includes:

[0039] Step 7.1: The accounting agency sends a carbon footprint data request to the upstream supplier through the system;

[0040] Step 7.2: The supplier system automatically returns the verifiable certificate of its product carbon footprint; if the supplier does not provide it, the accounting agency will communicate offline to obtain it;

[0041] Step 7.3: The accounting agency calls the blockchain service, passes in the credential ID and metadata, and verifies the authenticity and validity of the credential.

[0042] Specifically, the steps in step 9 where the accounting agency completes the carbon footprint accounting certificate and generates a verifiable carbon footprint voucher include:

[0043] Step 9.1: The accounting agency completes the carbon footprint accounting based on the product carbon footprint model and supply chain carbon emission data, and outputs the carbon footprint report;

[0044] Step 9.2: Convert the carbon footprint report into JSON-LD format that complies with the W3C Verifiable Credentials specification, mainly including:

[0045] Credential metadata: context, type, issuer, issueDate;

[0046] credentialSubject: contains information on calculation methods, emissions, and emission reduction recommendations;

[0047] Proof: Use the private key corresponding to the digital identity of the accounting institution to sign the certificate to ensure the integrity and non-repudiation of the certificate;

[0048] Step 9.3: The accounting agency sends the verifiable carbon footprint certificate to the manufacturing company and writes the certificate metadata into the blockchain, binding it to the company and product identity.

[0049] Specifically, the steps in step 10 whereby the production enterprise obtains the carbon footprint certificate and publicly verifies the traceability include:

[0050] Step 10.1: The manufacturer obtains verifiable carbon footprint credentials, which can be provided to downstream enterprises and consumers through public or shared means;

[0051] Step 10.2: The verifier obtains the carbon footprint certificate file or ID and verifies it through the API interface provided by the blockchain system or certificate verification service;

[0052] Step 10.3: The verification service verifies the authenticity of the voucher based on the signature and evidence information in the voucher, and returns the voucher content for the verifier to browse the accounting details;

[0053] Step 10.4: The verifier can trace the carbon footprint certificate of the upstream enterprise through the product and enterprise identity in the certificate to achieve carbon footprint traceability of the entire supply chain.

[0054] The present invention has the following beneficial effects:

[0055] 1. Integrate domain ontology to build product carbon footprint knowledge graph, extract key concepts and relationships of carbon emissions in product life cycle, form a shared knowledge base and give it deep semantic representation, provide unified language and standardized process for product carbon footprint accounting, alleviate the dependence of accounting process on practitioners’ experience, and reduce the complexity of manual data collection and calculation;

[0056] 2. Introducing low-dimensional semantic representation of knowledge graph into deep learning model. Compared with the prediction method based solely on time series emission data, the fusion of graph knowledge can more accurately capture the dynamic evolution characteristics of product carbon footprint, enhance the stability and interpretability of the prediction model, and provide quantitative support for carbon footprint scenario analysis and emission reduction strategy formulation;

[0057] 3. Based on blockchain, distributed digital identity, verifiable credentials and other technologies, a product carbon footprint data collaborative network is established among supply chain enterprises, supporting the encryption and storage of carbon emission data on the chain, authorized sharing, and ensuring data ownership, interoperability and tamper-proofing, and realizing the secure cross-enterprise circulation of carbon footprint data in the form of trusted credentials, thereby improving the carbon accounting efficiency of the entire supply chain;

[0058] 4. Using blockchain traceability, the carbon footprint data of each link of the product is stored in a chain-based certificate to form an inseparable carbon footprint evidence chain. Stakeholders only need to obtain the product DID to verify and trace its carbon footprint throughout its life cycle, and gain an in-depth understanding of the carbon emission contribution of upstream companies, providing decision-making support for the promotion of low-carbon products and green supply chain governance.

[0059] To sum up, the present invention innovatively integrates emerging technologies such as knowledge graphs, blockchains, and verifiable credentials from both theoretical and practical levels, systematically solving pain points such as data silos, knowledge fragmentation, and collaborative inefficiency in the field of product carbon footprint, and forming a new paradigm of carbon management that can be promoted and applied on a large scale. It is of great significance to promote the digital and green transformation of the industry and promote the implementation of the "dual carbon" goals. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flow chart of the supply chain carbon footprint tracing method of the digital identity and verifiable credentials of the present invention;

[0061] Figure 2 This is a schematic diagram of the framework of the supply chain carbon footprint traceability system of the digital identity and verifiable credentials of the present invention. DETAILED DESCRIPTION

[0062] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0063] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0064] The specific scheme of the real-time identification method of investment risks based on knowledge graph and anomaly detection provided by the present invention is described in detail below with reference to the accompanying drawings.

[0065] See also Figure 1 , a supply chain carbon footprint traceability method based on digital identity and verifiable credentials, including the following steps:

[0066] Step 1: Construct a product carbon footprint knowledge graph based on the ontology to obtain the entities, relationships, and constraints of each product life cycle stage.

[0067] Specifically, this step includes:

[0068] Step 1.1: Based on the OWL ontology description language, define the core concepts, attributes and relationships of each link in the product life cycle and build a TBox knowledge base. The core concepts include but are not limited to: products, raw materials, production, transportation, use, recycling, etc.; relationships include composition, emissions, impact, etc.; attributes include product name, model, weight, energy consumption, emission factor, etc.;

[0069] Step 1.2: Use natural language processing, data mining and other technologies to extract entities and relationships from structured and unstructured data to build an ABox assertion library. Structured data includes product BOM tables, emission factor databases, etc., and unstructured data includes product manuals, carbon footprint reports, etc. For example, extract composition relationships such as "battery packs contain positive electrode materials, negative electrode materials, electrolytes, and separators" from the product BOM table; extract emission data such as "carbon emissions during the battery pack production phase are 320kgCO2e / unit" from the carbon footprint report;

[0070] Step 1.3: Combine TBox and ABox to form the ALCQ(D) description logic knowledge base, and use JenaAPI to serialize it into an RDF format knowledge graph. At this point, the product carbon footprint knowledge graph is constructed, formally representing the carbon emission-related concepts, instances and their semantic associations at each life cycle stage of the product.

[0071] Step 2: Perform representation learning on the knowledge graph through the hierarchical relational attention network HRGAT to obtain a low-dimensional real-valued vector.

[0072] Specifically, this step includes:

[0073] Step 2.1: Randomly initialize the entity embedding vector u according to the knowledge graph G = (E, R) i ∈R d and the relation embedding vector v r ∈R d . Where E and R represent the entity set and relationship set in the knowledge graph, respectively, and d is the embedding vector dimension, which generally ranges from 100 to 500. The initial embedding vector follows the Xavier normal distribution, that is,

[0074] Step 2.2: For an entity type t∈T, compute the type embedding w by aggregating the entity embeddings belonging to t through attention t :

[0075]

[0076] Where T is the set of entity types, Et is the set of entities belonging to type t, σ is the ReLU activation function, and f(·) is the attention scoring function, which can be a vector dot product, matrix transformation, etc. is the normalized attention weight of entity i relative to type t;

[0077] Step 2.3: Iteratively update the entity embedding h through the hierarchical attention mechanism i :

[0078]

[0079] Where l is the number of HRGAT layers, which is generally 1 to 3; T i is the type set containing entity i; S i is the set of relationship types containing entity i; N r (i) is the set of neighbor entities connected to entity i through relationship r; is the attention weight of entity i and type t, reflecting the strength of association between entity i and the type: is the attention weight of entity j to entity i through relationship s, reflecting the interaction strength between the two entities under relationship s; W s′ b s is the transformation matrix and bias term corresponding to the relation s; Step 2.4: Repeat step 2.3 until the entity embedding converges or reaches the preset number of iterations, and output the final entity embedding representation {h i} and relation embedding representation {v r}. The embedding vector dimension d is generally set to 100 to 500, and the number of iterations is generally 20 to 100. At this point, the knowledge graph representation learning is completed.

[0080] Step 3: Integrate the knowledge graph embedding vector into the long short-term memory network LSTM to build the LKGC-LSTM carbon footprint prediction model. Specifically, this step includes:

[0081] Step 3.1: Transform the observation x at the i-th time step into i Mapped to LSTM input vector input i =W i x i +b i ∈R k Among them, x i It can be the carbon emission related data (energy consumption, output, etc.) of day / week / month i, W i ,b i is the mapping matrix and bias term, k is the input vector dimension, generally 50 to 200;

[0082] Step 3.2: Embed the knowledge graph into vector ui (embedded vector corresponding to product entity i) and input i Splice, get This step introduces product background knowledge into time-series carbon emission forecasting;

[0083] Step 3.3: Based on Perform LSTM state transfer:

[0084]

[0085] h i =o i tanh(C i )

[0086] Among them, f i 、i i , o i They are forget gate, input gate, and output gate respectively. is the candidate state, W f / W i / W o / W C and b f / b i / b o is the weight matrix and bias term corresponding to the gate. The hidden state dimension is generally 50 to 200. The gate activation function σ is the sigmoid function, and the state activation function is the hyperbolic tangent function tanh.

[0087] Step 3.4: Use MLP to transform the LSTM final hidden state h t Mapping for the future Among them, MLP consists of 1 to 3 fully connected layers, each with 50 to 200 neurons, and the activation function is ReLU. N is the prediction step size, which is generally 7, 14, 30, etc., corresponding to the prediction of the carbon footprint in the next week, two weeks, and one month. When training the model, the root mean square error (RMSE) is used as the loss function, and the Adam optimizer is used for training, with a learning rate of 0.001 to 0.01.

[0088] Step 4: The manufacturer registers an account and generates a unique digital identity, registers the product identity and associates it with the company identity. Specifically, this step includes:

[0089] Step 4.1: The enterprise submits registration information including enterprise name, unified social credit code, user name and password, etc. Among them, the unified social credit code is the only identity mark registered and publicized by the enterprise in the government department, which can be used to verify the authenticity of the enterprise;

[0090] Step 4.2: The system backend calls the enterprise credit information disclosure system API, inputs the enterprise name and unified social credit code, and obtains credit information such as enterprise registration information, business status, administrative penalties, etc. If the information matches and the enterprise status is normal, the application is approved; otherwise, an error prompt is returned and the user is required to resubmit the application.

[0091] Step 4.3: Generate a public key and digital identity based on the public and private keys in the did:key mode. Specifically, use the Secp256k1 elliptic curve algorithm to generate a pair of public and private keys (pk i ,sk i ), and use BASE58 to encode the public key to obtain the enterprise distributed digital identity did i . Private key sk i It is kept properly by the enterprise and used for identity authentication and data signature;

[0092] Step 4.4: Use blockchain smart contract to add the enterprise account address addr i 、Digital identity i 、Public key pk i The registration information is recorded in the blockchain account book in the form of transactions. Anyone can register with the digital identity i The registration information of the corresponding enterprise can be found, but it cannot be tampered with or forged;

[0093] Step 4.5: The enterprise submits basic product information such as name, model, etc., and the system calls the digital identity registration contract to generate the product digital identity did p The product identity also adopts the didikey model, and the generation process is similar to that of the corporate identity;

[0094] Step 4.6: Use IoT electronic tag technology such as NFC and RFID to generate a unique ID for each product entity and associate it with its digital identity p Binding. Specifically, the NFC tag stores the product ID and digital identity p The binding relationship is recorded on the blockchain through the IoT identity contract. Scan the product NFC tag to obtain the corresponding product DID document and view product information;

[0095] Step 4.7: Duplicate product digital identity p and corporate identity i The relationship between the product and the enterprise is registered on the blockchain through the product identity contract. In this way, the product DID can be traced back to the enterprise to which it belongs to establish a trusted mapping between the product and the enterprise identity.

[0096] Step 5: The manufacturer submits a carbon footprint audit application to the accounting agency.

[0097] Specifically, the manufacturing enterprise fills in and submits the carbon footprint accounting application through the management system, which mainly includes: enterprise identity information i , product identity information {did p}, annual carbon inventory, list of emission facilities, activity level data, etc. Accounting agency managers can view and accept applications after logging into the system.

[0098] Step 6: The accounting institution registers an account and generates a digital identity, accepts the audit application and conducts carbon footprint modeling and data collection. Specifically, the accounting institution submits the registration application, and the system verifies the institution's qualifications and generates the institution's digital identity through the did:key mode. o , and deploy the identity registration contract to record it on the blockchain. For the accepted carbon footprint audit applications, the accounting agency uses the management system to set carbon footprint boundaries, identify emission sources, collect activity level data, select emission factors, and form a product carbon footprint model.

[0099] Step 7: The accounting agency collects verifiable carbon footprint certificates from supply chain companies and calls the blockchain service to verify the authenticity of the certificates. Specifically, this step includes: Step 7.1: The accounting agency sends a carbon footprint data application to the upstream supplier through the system. The application content includes the supplier's corporate identity, relevant product identity, time range and accuracy of carbon emission data requirements, etc.; Step 7.2: After the supplier system receives the application, it automatically calls the carbon footprint certificate generation service, passes in relevant product identity, time range and other parameters, generates and returns a verifiable certificate that complies with the W3CVC specification. If the supplier is not connected to the certificate generation service, the accounting agency shall manually request and enter the supplier's carbon emission data; Step 7.3: The accounting agency calls the blockchain service and passes in the verifiable certificate VC returned by the supplier. s and metadata such as corporate identity s 、Product identity p etc., verify the authenticity and validity of the credentials. The verification process is as follows:

[0100] According to the certificate issuer field, obtain the supplier's enterprise public key pk s ;

[0101] Verify that the certificate signature is signed by the public key pk s The corresponding private key sk s Issue;

[0102] According to the enterprise identity in the credential did s 、Product identity v , query the blockchain ledger to verify the consistency between the credential subject and the declared content;

[0103] Determine whether the certificate is expired based on the certificate issuance time and validity period; if the certificate verification passes, the carbon emission data in it will be included in the product carbon footprint accounting; otherwise, output the reason for the verification failure and require the supplier to resubmit the certificate.

[0104] Step 8: The accounting agency integrates the LKGC-LSTM carbon footprint prediction model to complete the accounting and generate a carbon footprint report. Specifically, after completing the construction of the product carbon footprint model, direct emission data collection, and supply chain emission data collection, the accounting agency inputs various primary data into the LKGC-LSTM carbon footprint prediction model, and combines the product knowledge graph to predict the carbon emissions of the product within a certain time range in the future. Then, according to the product functional unit, the unit product carbon footprint is converted. Finally, referring to ISO / TS14067, PAS2050, GHGProtocol and other standards, a product carbon footprint report is compiled, which mainly includes: accounting purpose, result overview, data source, system boundary, allocation method, emission factor, uncertainty analysis and other chapters.

[0105] Step 9: The accounting agency completes the carbon footprint accounting certificate and generates a verifiable carbon footprint certificate after signing with a digital identity. Specifically, this step includes: Step 9.1: The accounting agency fills in the standardized carbon footprint accounting certificate based on the product carbon footprint report, including the product name, carbon footprint results, accounting basis, verification agency, certificate validity period, etc., and stamps it with the official seal;

[0106] Step 9.2: Convert the carbon footprint accounting certificate into JSON-LD format according to the W3CVC specification to form a verifiable certificate VC c The credential field contains the accounting certificate content, and the proof field contains the accounting institution's credential signature. The signing process is as follows: The accounting institution uses its own private key sk o Sign the content of the credential body to obtain the signature value Sig; add the signature value Sig to the proof field of the credential;

[0107] The final form of the proof field is as follows:

[0108] {"type":"Secp256k1Signature2020",

[0109] "created":"2021-06-23T17:13:12Z",

[0110] "verificationMethod":"did:key:z6Mkf...fKV#key-1",

[0111] "proofPurpose":"assertionMethod","proofValue":"Sig"}

[0112] Among them, type is the signature algorithm, created is the signature creation time, and verificationMethod is the verification institution digital identity did o and key reference, proofPurpose is the signature purpose, proofValue is the signature value;

[0113] Step 9.3: The accounting agency verifies the carbon footprint certificate VC through the blockchain verifiable certificate management contract c Associated to product identity did p , and write the certificate CID into the blockchain account book. Here CID stands for Content Identifier, which is the unique identifier of the IPFS network storage file. Anyone only needs the product DID to query and verify the carbon footprint certificate associated with it. Step 10: The manufacturer obtains the carbon footprint certificate and can disclose it to the verifier or upload it to the carbon disclosure platform for verification and traceability.

[0114] Specifically, this step includes:

[0115] Step 10.1: The manufacturer queries the list of carbon footprint certificates associated with its product DID from the blockchain verifiable certificate management contract;

[0116] Step 10.2: The manufacturer can share the carbon footprint certificate of a product with downstream enterprises, consumers and other stakeholders, or take the initiative to disclose it on the carbon information disclosure platform and accept supervision from all sectors of society;

[0117] Step 10.3: The Verifier obtains the carbon footprint certificate VC for a product c After that, it can be verified through the blockchain certificate verification service. The verification process is as follows: According to the certificate issuer field, obtain the accounting agency public key pk oi Verify whether the signature value Sig of the credential subject is generated by the public key pk o The corresponding private key sk o Issue; according to product identity did p , query the blockchain account book to verify whether the certificate is the latest and valid carbon footprint certificate of the product. If the certificate verification is successful, the certificate content will be returned to the verifier and the product carbon footprint accounting results will be presented; otherwise, a verification failure prompt will be displayed;

[0118] Step 10.4: The verifier can also use the blockchain traceability service to gain an in-depth understanding of the formation process of the product's carbon footprint and trace the carbon emission contribution of upstream suppliers. The traceability process is as follows: Based on the corporate identity in the product's carbon footprint certificate, i, obtain the manufacturing company information; traverse the bill of materials (BOM) in the product carbon footprint model to obtain the supplier identity, quantity, carbon emissions and other information of each component; call the certificate verification service to obtain the carbon footprint certificate of each supplier, verify the carbon emission data, and use the carbon footprint traceability diagram visualization technology to generate a carbon footprint tracking diagram containing suppliers at all levels, which intuitively shows the composition of the product carbon footprint.

[0119] Finally, see Figure 2 The present invention also provides a supply chain carbon footprint management system based on the above method, including: product carbon footprint knowledge graph construction module 1, knowledge graph representation learning module 2, carbon footprint prediction module 3, digital identity management module 4, carbon footprint certificate generation module 5, blockchain service module 6, etc. The modules exchange data and control instructions through the data bus to collaboratively complete product carbon footprint accounting, certification and full-process traceability management.

[0120] Among them, the knowledge graph construction module 1 is responsible for abstracting domain concepts, defining ontologies, integrating structured and unstructured data, and forming an extensible product carbon footprint knowledge base; the knowledge graph representation learning module 2 adopts a graph neural network based on an attention mechanism to learn the low-dimensional semantic representation of the knowledge graph; the carbon footprint prediction module 3 uses graph embedding and time-series emission data to train the product carbon footprint prediction model; the digital identity management module 4 performs distributed digital identity registration and management for entities such as enterprises, products, and accounting institutions, including DID creation, public and private key generation, identity contract deployment and calling, etc.; the carbon footprint certificate generation module 5 is responsible for generating a certificate file containing product carbon footprint information and a digital signature in accordance with the verifiable certificate standard, and registering it to the blockchain; the blockchain service module 6 uses a consortium chain platform such as Hyperledger Fabric to build a blockchain network environment for carbon footprint certificates and traceability, deploy certificate contracts and identity contracts, and provide API services such as on-chain data query and certificate verification.

[0121] It should be noted that the technical solution of the present invention is not limited to the above embodiments, and those skilled in the art can substantially make various modifications and variations, but these modifications and variations all belong to the protection scope of the present invention. For example, the construction of the product carbon footprint knowledge graph can adopt the ontology automation construction technology; the knowledge graph representation learning can adopt the random walk algorithm based on subgraph sampling; the LSTM network can be replaced by other time series prediction models such as GRU and Attention; the underlying blockchain platform can adopt public chains such as Ethereum.

[0122] The technical solutions of the present invention are described above based on specific embodiments, but these embodiments cannot limit the present invention to the specific implementation scope described. Any changes, combinations or substitutions made by any person skilled in the art to these features and embodiments without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined according to the claims.

Claims

1. A supply chain carbon footprint traceability method based on digital identity and verifiable credentials, characterized in that: It includes the following steps: Step 1: Construct a product carbon footprint knowledge graph based on the ontology to obtain entities, relationships and constraints at each product life cycle stage; Step 2: Perform representation learning on the knowledge graph through the hierarchical relational attention network HRGAT to obtain a low-dimensional real-valued vector; Step 3: Integrate the knowledge graph embedding vector into the long short-term memory network LSTM to construct the LKGC-LSTM carbon footprint prediction model; Step 4: The manufacturer registers an account and generates a unique digital identity, registers the product identity and associates the enterprise identity; Step 5: The manufacturer submits a carbon footprint audit application to the accounting agency; Step 6: The accounting agency registers an account and generates a digital identity, accepts the audit application and performs carbon footprint modeling and data collection; Step 7: The accounting agency collects verifiable carbon footprint credentials of supply chain enterprises and calls blockchain services to verify the authenticity of the credentials; Step 8: The accounting agency integrates the LKGC-LSTM carbon footprint prediction model to complete the accounting and generate a carbon footprint report; Step 9: The accounting agency completes the carbon footprint accounting certificate and generates a verifiable carbon footprint certificate after signing with a digital identity; Step 10: The production enterprise obtains the carbon footprint certificate and can disclose it to the verifier or upload it to the carbon disclosure platform for verification and traceability.

2. The method according to claim 1, characterized in that The steps of constructing a product carbon footprint knowledge graph based on ontology in step 1 include: step 1.1: based on the OWL ontology description language, define the core concepts, attributes and relationships of each link in the product life cycle, and build a TBox knowledge base; step 1.2: use natural language processing and data mining technology to extract entities and relationships from structured and unstructured data, and build an ABox assertion library; step 1.3: combine TBox and ABox to form an ALCQ (D) description logic knowledge base, and use JenaAPI to serialize it into an RDF format knowledge graph.

3. The method according to claim 1, characterized in that The step of performing representation learning on the knowledge graph by using HRGAT in step 2 includes: Step 2.1: Randomly initialize the entity embedding vector u according to the knowledge graph G = (E, R) i ∈R d and the relation embedding vector v r ∈R d ; Step 2.2: For an entity type t∈T, compute the type embedding w by aggregating the entity embeddings belonging to t through attention t : Among them, σ is the activation function, f(·) is the attention scoring function; similarly, the embedding w of the relation type s∈S can be obtained s ; Step 2.3: Iteratively update the entity embedding h through the hierarchical attention mechanism i : in, is the attention weight of entity i and type t; is the attention weight of entity j on entity i via relation s; Step 2.4: Repeat step 2.3 until the entity embedding converges and output the final embedding representation.

4. The method according to claim 1, characterized in that: The steps of constructing the LKGC-LSTM carbon footprint prediction model in step 3 include: Step 3.1: transforming the observation x at the i-th time step i Mapped to LSTM input vector input i =W i x i +b i ∈R k ; Step 3.2: Embed the knowledge graph into vector u i With input i Splice, get Step 3.3: Based on Perform LSTM state transfer: h i =o i fishy(C) i ) Among them, f i 、i i , o i They are forget gate, input gate, and output gate respectively. is the candidate state, W and b are the parameter matrix and bias term; Step 3.4: Use MLP to transform the LSTM final hidden state h t Mapped to the predicted carbon footprint value for the next N steps 5. The method according to claim 1, characterized in that The step of the manufacturing enterprise registering an account and generating a unique digital identity in step 4 includes: Step 4.1: The enterprise submits registration information including enterprise name, unified social credit code, username and password; Step 4.2: The system verifies the enterprise registration information and opens an account for the enterprise user after the application is approved; Step 4.3: Generate the enterprise digital identity based on the public and private keys through the did:key mode, and the public key and digital identity are equivalent; Step 4.4: Use blockchain smart contracts to register the enterprise account address and digital identity information to the distributed identity registration center.

6. The method according to claim 5, characterized in that The steps of registering the product identity and associating the enterprise identity in step 4 include: Step 4.5: The enterprise submits basic product information such as name and model, and the system generates a digital product identity; Step 4.6: Generate a unique ID for each product entity using the IoT electronic tag and bind it to its digital identity; Step 4.7: Register the association between the product digital identity and the enterprise identity to the blockchain identity registration center.

7. The method according to claim 1, characterized in that The steps in step 7 where the accounting agency collects verifiable carbon footprint certificates from supply chain companies include: Step 7.1: The accounting agency sends a carbon footprint data request to the upstream supplier through the system; Step 7.2: The supplier system automatically returns the verifiable certificate of its product carbon footprint; if the supplier does not provide it, the accounting agency will communicate offline to obtain it; Step 7.3: The accounting agency calls the blockchain service, passes in the credential ID and metadata, and verifies the authenticity and validity of the credential.

8. The method according to claim 1, characterized in that The steps in step 9 where the accounting agency completes the carbon footprint accounting certificate and generates a verifiable carbon footprint certificate include: Step 9.1: The accounting agency completes the carbon footprint accounting based on the product carbon footprint model and supply chain carbon emission data, and outputs the carbon footprint report; Step 9.2: Convert the carbon footprint report into JSON-LD format that complies with the W3C Verifiable Credentials specification, mainly including: Credential metadata: @context, type, issuer, issueDate; credentialSubject: contains information on calculation methods, emissions, and emission reduction recommendations; Proof: Use the private key corresponding to the digital identity of the accounting institution to sign the certificate to ensure the integrity and non-repudiation of the certificate; Step 9.3: The accounting agency sends the verifiable carbon footprint certificate to the manufacturing company and writes the certificate metadata into the blockchain, binding it to the company and product identity.

9. The method according to claim 1, characterized in that: The steps in step 10 for the production enterprise to obtain the carbon footprint certificate and publicly verify the traceability include: Step 10.1: The manufacturer obtains verifiable carbon footprint credentials, which can be provided to downstream enterprises and consumers through public or shared means; Step 10.2: The verifier obtains the carbon footprint certificate file or ID and verifies it through the API interface provided by the blockchain system or certificate verification service; Step 10.3: The verification service verifies the authenticity of the voucher based on the signature and evidence information in the voucher, and returns the voucher content for the verifier to browse the accounting details; Step 10.4: The verifier can trace the carbon footprint certificate of the upstream enterprise through the product and enterprise identity in the certificate to achieve carbon footprint traceability of the entire supply chain.

Citation Information

Cited By

  • Green product authentication and tracing system based on block chain

    CN120181875A

  • A blockchain-based green product certification and traceability system

    CN120181875B

  • Block chain-based product carbon data full-process management method and system

    CN120806351A

  • Battery carbon footprint determination method and device and storage medium

    CN121094341A

  • Carbon data evidence storage method and system in combination with block chain and cryptographic algorithm

    CN121508797A