Electronic waste recovery cross-chain authentication method and system based on multi-agent system

Abstract

The invention belongs to the field of waste electronic and electric appliance recovery authentication, and relates to an electronic waste recovery cross-chain authentication method and system based on a multi-agent system. The authentication system comprises an enterprise authentication submission agent, a cross-chain information integration agent, an authentication compliance analysis agent and a feedback optimization agent. The enterprise authentication submission agent is used for managing and verifying an initial data submission process of an enterprise; the cross-chain information integration agent is used for collecting, integrating and verifying cross-chain data; the authentication compliance analysis agent is used for integrating data and carrying out compliance analysis, judging whether an enterprise meets a GRS authentication standard or not and giving an audit result; the feedback optimization agent is used for analyzing the GRS audit report provided by the authentication compliance analysis agent and generating an improvement suggestion; according to the authentication method provided by the invention, the data accuracy, transparency and security of the GRS authentication process can be improved, and the efficiency and reliability of the enterprise authentication process are optimized.

Description

Technical Field

[0001] The present invention belongs to the field of blockchain technology and waste electronic appliance recycling certification, and relates to an electronic waste recycling cross-chain certification method and system based on a multi-agent system. Background Art

[0002] With the increasing awareness of global carbon emission reduction and the increasing importance of resource recycling, the recycling and reuse of electronic waste has become an important research direction. Electronic waste management is not only related to the compliance with environmental protection laws and regulations, but also directly affects the effective recycling of resources.

[0003] The entire process of recycling, disassembling and reusing waste electrical and electronic equipment (WEEE) involves multiple links and enterprises, which require efficient and transparent collaboration. In order to ensure the normalization and standardization of the entire recycling process, GRS (Global Recycling Standard) certification has become an important industry standard. GRS certification requires companies to meet certain environmental and social responsibility standards during production and recycling. However, traditional certification processes often rely on single on-chain information and manual audits, and there are problems such as information islands, inconsistent data, and difficulty in cross-chain integration. This limitation greatly affects the efficiency and accuracy of certification.

[0004] In this context, existing methods aim to improve the transparency and efficiency of the certification process in the field of e-waste recycling by utilizing blockchain technology. However, each enterprise in the field of e-waste recycling (recycling, disassembly, and remanufacturing) usually maintains an independent blockchain system, and data is difficult to share effectively between different enterprises. Due to the lack of a unified cross-chain information integration mechanism, the certification body cannot fully obtain the relevant data of each enterprise when conducting compliance verification, resulting in serious information island problems, which in turn affects the comprehensiveness and accuracy of the certification. Secondly, the ability to handle sensitive information is insufficient. In the existing certification process, the sensitivity of enterprise data is difficult to solve effectively. Manual analysis and processing of sensitive information is time-consuming and error-prone, and the security and compliance of corporate confidential information cannot be guaranteed. At the same time, the lack of automated sensitive information processing means leads to slow generation of certification reports, affecting the normal operation of enterprises. Finally, data consistency and cross-chain collaboration are difficult. In the existing certification process, the integration and verification of cross-chain data is a huge challenge. Since the blockchain systems of each enterprise operate independently, data is prone to delays or losses during cross-chain transmission, making it difficult to ensure data consistency and integrity. This not only increases the complexity of certification, but may also lead to deviations in compliance judgments, affecting the fairness and authority of GRS certification.

[0005] Therefore, how to effectively integrate and verify data between different enterprises and blockchains to ensure the fairness and accuracy of GRS certification has become an urgent problem to be solved. Summary of the invention

[0006] In response to the above technical problems, the present invention provides a cross-chain certification method for electronic waste recycling based on a multi-agent system to solve the problems of high complexity of cross-chain data integration, insecure processing of sensitive information, low efficiency of compliance analysis, and imperfect feedback mechanism in the existing system, thereby improving the data accuracy, transparency and security of the GRS certification process, and optimizing the efficiency and reliability of the enterprise certification process.

[0007] The present invention is achieved through the following technical solutions: An electronic waste recycling cross-chain certification system based on a multi-agent system, the electronic waste recycling cross-chain certification system comprising: an enterprise certification submission agent, a cross-chain information integration agent, a certification compliance analysis agent, and a feedback optimization agent; The enterprise certification submission agent is used to extract data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises, process the sensitive information of the enterprises, conduct preliminary compliance checks on the processed non-sensitive data, generate an initial declaration report and submit the initial declaration report to the cross-chain information integration agent; The cross-chain information integration agent is used to perform cross-chain retrieval, integration and transmission of data between blockchains of different enterprises to ensure the consistency, security and integrity of data in cross-chain operations; The certification compliance analysis agent is used to receive the integrated data packet sent by the cross-chain information integration agent, perform compliance analysis, determine whether the corresponding enterprise meets the GRS certification standards, and provide a GRS audit report; if the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the supervision chain; if the enterprise fails to pass the audit, the GRS audit report is transmitted to the feedback optimization agent; The feedback optimization agent is used to analyze the GRS audit report provided by the certification compliance analysis agent, generate improvement suggestions, and pass the improvement suggestions to the enterprise certification submission agent through the cross-chain information integration agent to guide the enterprise to make improvements.

[0008] Furthermore, the cross-chain information integration agent obtains the required authentication data from the blockchains of different enterprises; the authentication data includes the initial declaration report and the supply chain data of upstream and downstream enterprises; the supply chain data of upstream and downstream enterprises includes supplier compliance information and supply chain transparency information; supplier compliance information involves whether the supplier has relevant environmental certification and whether it complies with GRS certification requirements in the production and procurement links; supply chain transparency information refers to the traceability of each link of the entire supply chain, ensuring that the data of each link can be recorded and verified, and the source and processing process of raw materials can be clearly traced; After acquiring the data, the authentication compliance analysis agent integrates the data from different sources to ensure that all data remain synchronized; during the integration process, data consistency verification is performed to ensure that the information on each blockchain matches each other; the integrated data packet is stored on the relay blockchain after consensus, and then encrypted and transmitted to the authentication compliance analysis agent; Furthermore, the cross-chain information integration agent has the function of feedback transmission. When the feedback optimization agent generates improvement suggestions, the cross-chain information integration agent transmits the improvement suggestion feedback to the enterprise certification submission agent.

[0009] Furthermore, the certification compliance analysis agent uses a large language model to conduct an in-depth analysis of the integrated data packets sent by the cross-chain information integration agent to determine whether the operations of each enterprise comply with regulatory requirements and environmental protection standards, and generate a GRS audit report; If the company passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the chain of custody; if the company fails the audit, the GRS audit report is transmitted to the feedback optimization agent; Furthermore, the certification compliance analysis intelligent body conducts a secondary review of the data resubmitted by the enterprise after improvements.

[0010] An electronic waste recycling cross-chain authentication method based on a multi-agent system, the electronic waste recycling cross-chain authentication method comprising the following steps: S1: initialization of each agent and provision of local knowledge; S2: Data processing and initial declaration submission by the enterprise certification submission agent: The enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises, processes the sensitive information of the enterprises, performs a preliminary compliance check on the processed non-sensitive data, generates an initial declaration report and submits the initial declaration report to the cross-chain information integration agent; S3: The cross-chain information integration agent integrates the information submitted by each enterprise: the cross-chain information integration agent retrieves the required authentication data from the blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises; integrates and verifies the data on different blockchains. After obtaining the data, the authentication and compliance analysis agent integrates the data from different sources to ensure that all data remain synchronized; data consistency verification is performed during the integration process to ensure that the information on each blockchain matches each other; the integrated data packet is stored on the relay blockchain after PBFT consensus, and finally transmitted to the authentication and compliance analysis agent; S4: The certification compliance agent generates an analysis report based on the GRS certification standards: The certification compliance analysis agent uses a large language model to conduct an in-depth analysis of the integrated data packet sent by the cross-chain information integration agent to determine whether the operations of each enterprise comply with regulatory requirements and environmental protection standards, and generate a GRS audit report. The GRS audit report shows whether the enterprise has passed the GRS certification; If the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the chain of custody to ensure that the certification results are open, transparent and cannot be tampered with, and then proceeds to step S6; If the enterprise fails the audit, the GRS audit report is transmitted to the feedback optimization agent and proceeds to step S5; S5: Feedback optimization agent analyzes audit results and proposes enterprise improvement suggestions: The feedback optimization agent receives and analyzes the GRS audit report provided by the certification compliance agent, generates improvement suggestions, and transmits the improvement suggestions to the enterprise certification submission agent through the cross-chain information integration agent to guide the enterprise to make improvements; S6: If the enterprise passes the certification, the certification process ends; If the enterprise certification fails, the improvement suggestions are passed to the enterprise certification submission agent through the cross-chain information integration agent, and steps S2 to S5 are repeated until the enterprise certification passes.

[0011] Further, step S1 includes: Agent initialization: allocate the required computing resources and storage space for the enterprise certification submission agent, cross-chain information integration agent, certification compliance analysis agent, and feedback optimization agent; initialize the algorithm models inside the agent, such as the natural language processing module based on the large language model and the blockchain data retrieval module; set the communication protocol and data format between agents to ensure smooth and accurate information transmission; Local knowledge provision: Provide each agent with the necessary local knowledge base, which includes enterprise-related regulations, standard operating procedures, and historical certification data.

[0012] Further, step S2 includes the following steps: Step 2.1: Data retrieval and processing: The enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises. The data related to GRS certification includes production records and environmental compliance data; After extracting the data, the enterprise certification submission agent uses a large language model to conduct an in-depth analysis of the data and identify key information in the data; specifically, the key information includes information related to production records and information related to environmental compliance. Among them, information related to production records includes product source information, production process information, product batch information, and product quality information; information related to environmental compliance includes emission records, energy usage, waste disposal records, and environmental audit reports.

[0013] Step 2.2: Sensitive information processing: The enterprise authentication submission agent analyzes and detects sensitive data, which includes information involving corporate business secrets or personal privacy; the enterprise authentication submission agent encrypts or specially marks the sensitive data to prevent unauthorized access and leakage risks; Step 2.3: Preliminary compliance analysis: The enterprise certification submission agent conducts a preliminary compliance check on the processed non-sensitive data, compares the enterprise's operating procedures and records with the GRS certification requirements, and verifies whether the enterprise's actual operations comply with relevant standards and regulations; Step 2.4: Initial declaration submission: The enterprise certification submission agent organizes the data processed by the preliminary compliance analysis in Step 2.3 into a formatted initial declaration report, submits the initial declaration report to the cross-chain information integration agent, and waits for feedback or further processing instructions.

[0014] Further, step S3 includes the following steps: Step 3.1: Cross-chain data collection: The cross-chain information integration agent establishes a connection with the blockchain system of each enterprise; through the cross-chain protocol, the authentication data is collected from the blockchain of each enterprise, and the authentication data includes the initial declaration report and the supply chain data of upstream and downstream enterprises; The mathematical expression of the cross-chain data collection process is as follows: ; in, Indicates the enterprise i In time t The submitted data set, Indicates the enterprise i In time t Submitted j Class data, data from different companies' blockchains Extracted from Indicates i The blockchain of an enterprise n Indicates the number of enterprises; During the data collection process, the cross-chain information integration agent uses data verification technology to confirm the integrity and accuracy of the data, and verifies that the data has not been tampered with or lost through hash verification or digital signature means; the mathematical description is: ; in, Representation data D The hash value of Indicates the enterprise received by the cross-chain information integration agent i In time t Submitted j Class data, Representation data D Timestamp, indicating the time when the data was submitted; Step 3.2: Data integration and synchronization: During the integration process, the cross-chain information integration agent will perform format conversion and field mapping on various types of data to ensure that the format and content of all data remain consistent; and the cross-chain information integration agent will perform data synchronization checks and compare the timestamps of the data submitted by each enterprise; if time differences or data conflicts are found, an automated data conflict resolution process will be performed to ensure that the final integrated data set is synchronized and conflict-free; the automated data conflict resolution process mainly handles conflicts through timestamp priority strategy and data verification. First, the system compares the timestamps of different data and gives priority to the latest data; if the timestamps are the same, the system uses hash verification and digital signature technology to confirm the integrity of the data to ensure that it has not been tampered with or lost. In the case where data cannot be selected directly, the system will try to merge data from different sources through data fusion by analyzing the differences between different data items to generate a conflict-free, unified record.

[0015] The automated data conflict resolution process handles conflicts through a timestamp priority strategy and data verification, including the following steps: Compare the timestamps of different data and select the latest data, as shown below: ; in, For final submission of data, Representation data D timestamp; if the timestamps are the same, use hash check to confirm the integrity of the data to ensure that it has not been tampered with or lost; when data cannot be directly selected, data fusion is used to handle conflicts. The data fusion method is: Assume that the data fusion function is ; in, and are two conflicting data items, To provide a unified record without conflicts; For conflicting data of numerical type, the average value is used as the fused data; for conflicting data of string type, the result of string concatenation is used as the fused data, which is expressed as follows: ; Among them, Concatenate means the operation of concatenating two strings into one string; To integrate conflict-free records, use a merge operation: ; generate Final submission data, including m Indicates the total number of unified record items without conflict, i is an index, indicating the i conflict-free records, index i From 1 to m , traverse all conflict-free unified records; Step 3.3: Verification, consensus and transmission of integrated data: After completing data integration and synchronization, the final integrated data packet is verified to confirm the integrity and accuracy of all data; the data is stored in the relay blockchain after consensus using the PBFT consensus algorithm to ensure the security and traceability of the submitted data; finally, the verified and encrypted integrated data packet is transmitted to the certified compliance analysis agent.

[0016] Furthermore, in step 3.3, the data is stored in the relay blockchain after consensus is performed using the PBFT consensus algorithm, including the following steps: Step 1: PBFT consensus system initialization: Set the PBFT consensus system to contain n nodes, and allow at most f nodes to be malicious nodes; require n ≥ 3f + 1, and the nodes are divided into primary nodes and backup nodes; select a node as the primary node through rotation or random algorithm, and as long as the primary node does not switch, it is called a view; Step 2: Master node broadcast: In a certain view, the client sends the request <request, operation, timestamp, client> to the master node. If the client sends it to a slave node, the slave node forwards the request to the master node. The master node is responsible for broadcasting the request to all other replica nodes and completing consensus. Step 3: Response processing: All nodes complete the request processing and return the processing result <reply, view, timestamp, client, node id, response result> to the client; when the client checks and receives at least f+1 identical results from different nodes, it takes the identical results as the final result.

[0017] Further, step S4 includes the following steps: Step 4.1: Data analysis preparation: The authentication compliance analysis agent receives the integrated data packet from the cross-chain information integration agent and pre-processes the data; The preprocessing includes: structuring and formatting the data packet to ensure that the data meets the analysis requirements; checking whether the data packet contains all the necessary information, including the company's production records, environmental compliance documents, and supply chain traceability data; and conducting a preliminary check on the data's completeness and accuracy; Step 4.2: Certification compliance analysis: After data preprocessing is completed, the certification compliance analysis agent uses a large language model to conduct in-depth analysis of the data, determine whether the operations and records of each enterprise comply with the GRS certification standards, and generate analysis reports and certification recommendations; Step 4.3: Certification results and human intervention: During the certification compliance analysis process, human experts are allowed to intervene to review and verify the analysis report and certification recommendations of the certification compliance analysis agent. Based on the expert's review opinion, the certification compliance analysis agent adjusts the analysis results or modifies the report content to give the final audit results. Step 4.4: Final report generation: After confirming the final audit results, the certification compliance analysis agent generates the company's GRS audit report; if the company passes all certification requirements, a formal GRS certification certificate is generated and the GRS certification certificate is recorded on the regulatory blockchain; if the company fails the audit, the GRS audit report is transmitted to the feedback optimization agent.

[0018] Further, step S5 includes the following steps: Step 5.1: Receiving and analyzing audit results: The feedback optimization agent receives the GRS audit report provided by the certification compliance analysis agent; the feedback optimization agent analyzes the GRS audit report and identifies the weak links in the enterprise's certification process; the weak links include irregularities in the operating process, incomplete records, and substandard data quality; Step 5.2: Improvement suggestion generation: After analyzing the GRS audit report, the feedback optimization agent uses the large language model to generate improvement suggestions for the enterprise; Step 5.3: Suggestion transmission: After generating improvement suggestions, the feedback optimization agent transmits the improvement suggestions to the cross-chain information integration agent; the cross-chain information integration agent transmits the improvement suggestions to the relevant enterprise certification submission agent; the enterprise makes improvements and adjustments based on the improvement suggestions.

[0019] Beneficial technical effects of the present invention: The cross-chain authentication method for electronic waste recycling provided by the present invention realizes efficient data retrieval and integration between multiple blockchains based on a multi-agent system, greatly improving the efficiency of cross-chain data integration and reducing the complexity and time cost of manual intervention. At the same time, through the sensitive information processing mechanism, the security of sensitive information of enterprises during data submission is ensured, and the risk of information leakage is reduced. The use of an automated compliance analysis tool based on a large language model can accurately and efficiently verify the compliance of data submitted by enterprises, reduce the subjective bias of manual review, and thus improve the accuracy and consistency of authentication.

[0020] The present invention provides enterprises with detailed improvement suggestions through feedback optimization agents, and realizes closed-loop feedback through cross-chain information integration, helping enterprises to continuously optimize their processes and ensure the smooth progress of subsequent certification.

[0021] The cross-chain certification method for electronic waste recycling provided by the present invention not only enhances the transparency of the certification process, but also improves the credibility of the certification results, and significantly improves the credibility and efficiency of the entire GRS certification process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a schematic diagram of the architecture of the GRS cross-chain authentication system for electronic waste recycling enterprises based on multi-agents in an embodiment of the present invention; Figure 2 This is a flow chart of the GRS cross-chain authentication method for electronic waste recycling enterprises based on multi-agents in an embodiment of the present invention; Figure 3 This is a workflow diagram of an intelligent agent submitted for enterprise authentication in an embodiment of the present invention; Figure 4 This is a workflow diagram of a cross-chain information integration agent in an embodiment of the present invention; Figure 5 This is a flowchart of the PBFT formula algorithm used in the embodiment of the present invention; Figure 6 This is a workflow diagram of the authentication compliance analysis agent in an embodiment of the present invention; Figure 7 This is a workflow diagram of the feedback optimization agent in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] On the contrary, the present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention as defined by the claims. Further, in order to make the public have a better understanding of the present invention, some specific details are described in detail in the detailed description of the present invention below. Those skilled in the art can fully understand the present invention without the description of these details.

[0025] Example 1: Cross-chain certification system for electronic waste recycling based on a multi-agent system, such as Figure 1 As shown, the electronic waste recycling cross-chain certification system includes: an enterprise certification submission agent, a cross-chain information integration agent, a certification compliance analysis agent, and a feedback optimization agent; The enterprise certification submission agent is used to manage and verify the GRS certification process of each enterprise; the enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises, processes the sensitive information of the enterprises, performs a preliminary compliance check on the processed non-sensitive data, generates an initial declaration report and submits the initial declaration report to the cross-chain information integration agent; The cross-chain information integration agent is used to perform cross-chain retrieval, integration and transmission of data between blockchains of different enterprises to ensure the consistency, security and integrity of data in cross-chain operations; the cross-chain information integration agent is a bridge between the blockchains and regulatory chains of each enterprise to ensure seamless flow and accuracy of information.

[0026] The certification compliance analysis agent is used to receive the integrated data packet sent by the cross-chain information integration agent, perform compliance analysis, determine whether the corresponding enterprise meets the GRS certification standards, and provide a GRS audit report; if the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the supervision chain; if the enterprise fails to pass the audit, the GRS audit report is transmitted to the feedback optimization agent; The certification compliance analysis agent allows human intervention in the process of proposing certification result recommendations to give the final audit results. The combination of human and agent ensures the comprehensiveness and accuracy of the audit; as a key agent to ensure that the company complies with the GRS certification standards, the certification compliance analysis agent is not only responsible for analyzing the company's compliance status, but also undertakes the task of generating the final GRS certification certificate to ensure the authority and reliability of the certification process.

[0027] The feedback optimization agent is used to analyze the GRS audit report provided by the certification compliance analysis agent, generate improvement suggestions, and pass the improvement suggestions to the enterprise certification submission agent through the cross-chain information integration agent to guide the enterprise to make improvements.

[0028] In the present invention, a multi-agent system is a system composed of multiple agents, such as the AutoGen multi-agent framework, each of which is built on a large language model, and these agents can communicate, cooperate or compete with each other to complete specific tasks or goals. The AutoGen multi-agent framework is an open source AI framework released by Microsoft, which allows developers to use multiple agents to develop large language model applications. These agents can talk to each other to solve tasks. Each agent usually has autonomy, learning ability and interaction ability. The large language model is an artificial intelligence model based on deep learning that can understand and generate natural language, such as the ChatGPT language model. They are trained with a large amount of text data and have powerful language processing capabilities, and can perform tasks such as text generation, translation, and question-answering. The large language model understands and generates language in a wide range of contexts and can be used to provide complex reasoning and decision support. Specifically, the ChatGPT language model is an advanced language model based on the GPT architecture developed by OpenAI. Its full name is "ChatGenerative Pre-trained Transformer", which can generate natural and fluent dialogues and handle multiple language tasks by training with large-scale text data. The large language model and multi-agent system used in the present invention belong to conventional technologies.

[0029] In this embodiment, the enterprise certification submission agent is an independent agent distributed in each electronic waste recycling-related enterprise (such as recycling enterprises, dismantling enterprises, and remanufacturing enterprises). The main responsibility of the enterprise certification submission agent is to extract data related to GRS certification from the blockchain system to which the enterprise belongs. The data related to GRS certification include the production records and environmental compliance of the enterprise. After the data is extracted, the enterprise certification submission agent uses a large language model (LLM) to analyze the data to identify sensitive information. For the identified sensitive data, the enterprise certification submission agent adopts encryption or marking to protect it to ensure that the company's business secrets or other important information will not be leaked in the subsequent certification process. Next, the enterprise certification submission agent conducts a preliminary compliance analysis on non-sensitive data to determine whether the company's operating procedures meet the standard requirements of GRS. If the data meets the requirements, the enterprise certification submission agent will generate a preliminary GRS certification report. The GRS certification report covers the various indicators shown by the enterprise during the certification process and sends it to the cross-chain information integration agent for further processing and verification. In addition, the enterprise certification submission agent also has the function of feedback and improvement. When receiving improvement suggestions returned by the feedback optimization agent, the GRS certification report guides the enterprise to make necessary adjustments, such as re-extracting data or optimizing operating procedures, generating an updated certification report, and re-entering the certification process.

[0030] In this embodiment, the cross-chain information integration agent is a key agent deployed on the relay chain, which is mainly responsible for cross-chain retrieval, integration and transmission of data between blockchains belonging to different enterprises. Since electronic waste recycling involves multiple enterprises, and these enterprises may each maintain an independent blockchain system, the role of the cross-chain information integration agent is particularly important.

[0031] First, the cross-chain information integration intelligent body obtains the required authentication data from the blockchains of different enterprises; the authentication data includes the initial declaration report and the supply chain data of upstream and downstream enterprises; the supply chain data of upstream and downstream enterprises includes supplier compliance information and supply chain transparency information; supplier compliance information involves whether the supplier has relevant environmental certification, and whether it complies with GRS certification requirements in the production and procurement links; supply chain transparency information refers to the traceability of each link in the entire supply chain, ensuring that the data of each link can be recorded and verified, and the source and processing process of raw materials can be clearly tracked. These data are often distributed on multiple blockchains, so the cross-chain information integration agent is required to have strong cross-chain query capabilities; after obtaining the data, the certification compliance analysis agent integrates the data from different sources to ensure that all data remains synchronized and avoid compliance issues caused by data differences or delays; data consistency verification is performed during the integration process to ensure that the information on each blockchain matches each other and eliminate potential conflicts or errors; the integrated data packet is transmitted to the certification compliance analysis agent for further compliance review and certification; and the cross-chain information integration agent has the function of feedback transmission. When the feedback optimization agent generates improvement suggestions, the cross-chain information integration agent transmits the improvement suggestion feedback to the enterprise certification submission agent to help the enterprise improve its processes or data and ensure the continuity and consistency of certification.

[0032] In this embodiment, the certification compliance analysis agent is the core agent deployed on the supervision chain, responsible for detailed compliance analysis of the integrated data transmitted by the cross-chain information integration agent. The main task of the certification compliance analysis agent is to ensure that the operations and data of the enterprise comply with the GRS certification standards. The certification compliance analysis agent uses a large language model (LLM) to conduct an in-depth analysis of the integrated data packet sent by the cross-chain information integration agent to determine whether the operations of each enterprise comply with regulatory requirements and environmental protection standards. This analysis process not only involves the verification of the integrity and authenticity of the data, but also includes the compliance assessment of the enterprise's operating procedures. After completing the analysis, the agent will generate a detailed GRS audit report; this report will show whether the enterprise has passed the GRS certification and point out any deficiencies or problems in the certification process. If the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it on the supervision chain to ensure that the certification results are open, transparent and tamper-proof; if the enterprise fails to pass the audit, the GRS audit report is transmitted to the feedback optimization agent to generate specific improvement suggestions to help the enterprise make necessary adjustments and re-certification. In addition, the certification compliance analysis agent is also responsible for conducting a secondary review of the data resubmitted by the enterprise after improvement to ensure that it can ultimately comply with the GRS standards and obtain certification.

[0033] In this embodiment, the feedback optimization agent is an agent deployed on the supervision chain, which is specifically responsible for analyzing the audit report provided by the certification compliance analysis agent and generating specific improvement suggestions. Its role in the certification process is to ensure that the enterprise can meet the GRS certification standards through reasonable adjustments. The feedback optimization agent will first receive the audit results from the certification compliance analysis agent and conduct an in-depth analysis to identify the deficiencies of the enterprise in the certification process. These deficiencies involve the accuracy of the data and the compliance report of the operating process. Based on these analyses, the feedback optimization agent will generate detailed improvement suggestions, covering the specific links that the enterprise needs to adjust, and provide corresponding guidance to help the enterprise better understand the problem and make improvements. These suggestions usually include how to improve data processing methods and how to optimize operating procedures to meet the certification standards. After generating the suggestions, the feedback optimization agent will pass them to the enterprise certification submission agent through the cross-chain information integration agent, and the enterprise certification submission agent is responsible for guiding the enterprise to make improvements. Through this feedback mechanism, the feedback optimization agent helps enterprises continuously optimize their operations and data during the certification process, so as to ultimately meet the requirements of GRS certification and ensure the efficiency and continuity of the certification process.

[0034] Example 2: A cross-chain authentication method for electronic waste recycling based on a multi-agent system, which adopts the cross-chain authentication system for electronic waste recycling described in Example 1. The purpose of the cross-chain authentication method for electronic waste recycling is to use a multi-agent framework based on a large language model to solve the problems of high complexity of cross-chain data integration, insecure processing of sensitive information, low efficiency of compliance analysis, and imperfect feedback mechanism in the existing system, thereby improving the data accuracy, transparency and security of the GRS authentication process, and optimizing the efficiency and reliability of the enterprise authentication process.

[0035] like Figure 2 As shown, the electronic waste recycling cross-chain certification method includes the following steps: S1: initialization of each agent and provision of local knowledge, and step S1 also includes the role definition of each agent; S2: Data processing and initial declaration submission by the enterprise certification submission agent: The enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises, processes the sensitive information of the enterprises, performs a preliminary compliance check on the processed non-sensitive data, generates an initial declaration report and submits the initial declaration report to the cross-chain information integration agent; S3: The cross-chain information integration agent integrates the information submitted by each enterprise: the cross-chain information integration agent retrieves the required authentication data from the blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises; integrates and verifies the data on different blockchains. After obtaining the data, the authentication and compliance analysis agent integrates the data from different sources to ensure that all data remain synchronized; during the integration process, data consistency verification is performed to ensure that the information on each blockchain matches each other; the integrated data packet is transmitted to the authentication and compliance analysis agent; S4: The certification compliance agent generates an analysis report based on the GRS certification standards: The certification compliance analysis agent uses a large language model to conduct an in-depth analysis of the integrated data packet sent by the cross-chain information integration agent to determine whether the operations of each enterprise comply with regulatory requirements and environmental protection standards, and generate a GRS audit report. The GRS audit report shows whether the enterprise has passed the GRS certification; If the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the chain of custody to ensure that the certification results are open, transparent and cannot be tampered with, and then proceeds to step S6; If the enterprise fails the audit, the GRS audit report is transmitted to the feedback optimization agent and proceeds to step S5; S5: Feedback optimization agent analyzes audit results and proposes enterprise improvement suggestions: The feedback optimization agent receives and analyzes the GRS audit report provided by the certification compliance agent, generates improvement suggestions, and transmits the improvement suggestions to the enterprise certification submission agent through the cross-chain information integration agent to guide the enterprise to make improvements; S6: If the enterprise passes the certification, the certification process ends; If the enterprise certification fails, the improvement suggestions are passed to the enterprise certification submission agent through the cross-chain information integration agent, and steps S2 to S5 are repeated until the enterprise certification passes.

[0036] In this embodiment, step S1 includes the following steps: Step 1: Agent role definition. Role definition needs to be based on the needs of cross-chain certification of e-waste recycling of enterprises, and clarify the role played by each agent in the entire certification process. Specifically, the following responsibility definition is designed for the enterprise certification submission agent: "The enterprise certification submission agent is responsible for managing and verifying the initial data submission process of the enterprise"; the following responsibility definition is designed for the cross-chain information integration agent: "The cross-chain information integration agent is responsible for the collection, integration and verification of cross-chain data to ensure the consistency and integrity of data on different chains"; the following responsibility definition is designed for the certification compliance analysis agent: "The certification compliance analysis agent is responsible for receiving integrated data and conducting compliance analysis, determining whether the enterprise meets the GRS certification standards, and giving audit results"; the following responsibility definition is designed for the feedback optimization agent: "The feedback optimization agent is responsible for analyzing the audit results, generating improvement suggestions and guiding the enterprise to improve in order to increase the certification pass rate"; Step 2: Agent initialization. In the agent initialization phase, sufficient computing resources and storage space are allocated to each agent to ensure that it can efficiently perform complex tasks. Next, various algorithm models within the agent are initialized, including a natural language processing module based on a large language model and a blockchain data retrieval module to ensure that the agent has the ability to process and analyze multi-source data. Among them, the enterprise certification submission agent, the cross-chain information integration agent, the certification compliance analysis agent, and the feedback optimization agent are all equipped with a natural language processing module based on a large language model. In addition, the enterprise certification submission agent and the cross-chain information integration agent are also equipped with a blockchain data retrieval module. Finally, the communication protocol and data format between agents are configured to ensure that each agent can seamlessly connect during the data exchange process, achieve smooth and accurate information transmission, and thus ensure efficient collaboration of the entire system.

[0037] The natural language processing module based on the large language model is mainly used to parse and process text data from different sources, including certification materials, audit reports, production records, etc. submitted by enterprises. Through the powerful language understanding ability of the large language model, the module can identify and extract key information in the text, and automatically analyze the semantics of the data to ensure that all types of text data of the enterprise meet the requirements of GRS certification. The blockchain data retrieval module automatically completes data query and extraction by interacting with the smart contract on the blockchain. The data query smart contract is a code deployed on the blockchain to perform predefined data retrieval tasks. When the enterprise certification submission agent issues a request, the smart contract will automatically extract the corresponding production records and environmental compliance data information from the relevant blockchain (such as the blockchain of recycling enterprises, dismantling enterprises and remanufacturing enterprises). This module ensures that the extracted data has not been tampered with and meets the requirements of certification.

[0038] Step 3: Local knowledge provision. In the local knowledge provision stage, the necessary local knowledge base is distributed to each agent, covering enterprise-related regulations, standard operating procedures, and historical certification data, etc., to ensure that the agent has sufficient background information and reference basis when performing tasks. At the same time, it is ensured that each agent can access the latest GRS certification standards and related regulatory documents at any time to maintain the accuracy and compliance of its analysis and decision-making, and provide reliable support for subsequent certification and feedback processes.

[0039] In this embodiment, step S2 includes the following steps (such as Figure 3 shown): Step 2.1: Data retrieval and processing: The enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises. The data related to GRS certification include production records and environmental compliance data. After extracting the data, the enterprise certification submission agent uses a large language model to conduct in-depth analysis of the data and identify key information in the data. Specifically, the key information includes information related to production records and information related to environmental compliance. Among them, the information related to production records includes product source information (the source of raw materials of the product, whether it is recycled materials, and the composition analysis of the materials), production process information (whether environmental protection standards are followed during the production process, and whether the process used meets the requirements of sustainable development), product batch information (batch number, production time, and production location of the product), and product quality information (whether the product meets the proportion of recycled materials required by GRS certification); information related to environmental compliance includes emission records (waste gas, wastewater, and solid waste emission data during the production process of the enterprise), energy use (energy types used by the enterprise in the production process and their consumption), waste treatment records (treatment methods of waste generated during the production process), and environmental audit reports (whether the enterprise has received regular environmental compliance audits).

[0040] Step 2.2: Sensitive information processing: The enterprise authentication submission agent analyzes and detects sensitive data, which includes information involving corporate trade secrets or personal privacy. In order to ensure information security and privacy protection, the enterprise authentication submission agent encrypts or specially marks the sensitive data to prevent unauthorized access and leakage risks. After Step 2.2, all data are standardized and securely stored, ready to enter the next analysis stage.

[0041] Step 2.3: Preliminary compliance analysis: The enterprise certification submission agent conducts a preliminary compliance check on the processed non-sensitive data, compares the enterprise's operating procedures and records with the GRS certification requirements, and verifies whether the enterprise's actual operations comply with relevant standards and regulations; if the data is found to be incomplete or does not meet the certification standards during the analysis process, the agent will record these issues in detail and generate preliminary improvement suggestions based on the detected deficiencies. This step aims to ensure that the enterprise has a certain degree of compliance before formal certification through preliminary review, reducing problems and obstacles in subsequent audits.

[0042] Step 2.4: Initial declaration submission: The enterprise certification submission agent organizes the data after the preliminary compliance analysis and processing in Step 2.3 into a formatted initial declaration report. The enterprise certification submission agent organizes the data after the preliminary compliance analysis and processing in Step 2.3 into a formatted initial declaration report, submits the initial declaration report to the cross-chain information integration agent, and waits for feedback or further processing instructions.

[0043] In this embodiment, step S3 includes the following steps (eg Figure 4 shown): Step 3.1: Cross-chain data collection: The cross-chain information integration intelligent body establishes a connection with the blockchain system of each enterprise; through the cross-chain protocol, the authentication data is collected from the blockchain of each enterprise, and the authentication data includes the initial declaration report and the supply chain data of upstream and downstream enterprises; the supply chain data of upstream and downstream enterprises includes supplier compliance information and supply chain transparency information; supplier compliance information involves whether the supplier has relevant environmental certification, and whether it complies with GRS certification requirements in the production and procurement links; supply chain transparency information refers to the traceability of each link in the entire supply chain, ensuring that the data of each link can be recorded and verified, and the source and processing process of raw materials can be clearly tracked.

[0044] The mathematical expression of the cross-chain data collection process is as follows: ; in, Indicates the enterprise i In timet The submitted data set, Indicates the enterprise i In time t Submitted j Class data, data from different companies' blockchains Extracted from Indicates i The blockchain of an enterprise n Indicates the number of enterprises; During the data collection process, the cross-chain information integration agent uses data verification technology to confirm the integrity and accuracy of the data, and verifies that the data has not been tampered with or lost through hash verification or digital signature means; specifically, the mathematical description of the process is: ; in, Representation data D The hash value of Indicates the enterprise received by the cross-chain information integration agent i In time t Submitted j Class data, Representation data D The goal of this step is to ensure that all necessary information has been successfully extracted from each blockchain with high credibility.

[0045] Step 3.2: Data integration and synchronization: After successfully collecting blockchain data from different companies, the agent will integrate and standardize the data; during the integration process, the cross-chain information integration agent will convert the format and field mapping of various types of data to ensure that the format and content of all data remain consistent; and the cross-chain information integration agent will perform data synchronization checks, compare the timestamps of the data submitted by each company, and ensure that there are no obvious differences in the time of data submission; if time differences or data conflicts are found, an automated data conflict resolution process will be performed to ensure that the final integrated data set is synchronized and conflict-free, providing an accurate basis for subsequent analysis. Specifically, the above-mentioned automated data conflict resolution process mainly handles conflicts through timestamp priority strategies and data verification. First, the system compares the timestamps of different data and gives priority to the latest data, as shown below: ; in, For final submission of data, Representation data D timestamp; if the timestamps are the same, the system uses hash check to confirm the integrity of the data to ensure that it has not been tampered with or lost. If the data cannot be selected directly, the system will try to merge the data.

[0046] Specifically, assuming that the data fusion function is ; This function generates a conflict-free unified record by analyzing the differences between different data items. .

[0047] in, and are two conflicting data items, To provide a unified record without conflicts; Specifically, for conflicting data of numerical type, the average value is used as the fusion data; for conflicting data of string type, the result of string concatenation is used as the fusion data; the specific data fusion function is expressed as follows: ; Among them, Concatenate means the operation of concatenating two strings into one string; Next, the conflict-free unified records are integrated through the merge operation. ;generate Final submission data, including m Indicates the total number of unified record items without conflict, i is an index, indicating the i conflict-free records, index i From 1 to m , traverse all conflict-free unified records; Step 3.3: Integrated data verification and transmission: After completing data integration and synchronization, the final integrated data packet is verified to confirm the integrity and accuracy of all data; next, the PBFT consensus algorithm is used to reach data consensus and then stored in the relay blockchain to ensure the security and traceability of the submitted data; finally, the encrypted integrated data packet is transmitted to the certified compliance analysis agent.

[0048] Specifically, the PBFT consensus algorithm process is as follows: Figure 5 As shown, the overall basic process is as follows: Step 1: PBFT consensus system initialization: The PBFT consensus system is set to contain n nodes, of which at most f nodes are allowed to be malicious nodes. It is required that n ≥ 3f + 1. Nodes are divided into primary nodes and backup nodes. A node is selected as the primary node through rotation or random algorithm. As long as the primary node does not switch, it is called a view.

[0049] Step 2: Master node broadcast. In a certain view, the client will request<REQUEST, operation,timestamp,client> (ie <request, operation, timestamp, client>) is sent to the master node (if the client sends it to the slave node, the slave node will forward it to the master node), and the master node is responsible for broadcasting the request to all other replica nodes and completing the consensus.

[0050] Step 3: Response processing. All nodes complete the request processing and send the processing results.<REPLLY, view,timestamp, client, id_node, response> (ie <reply, view, timestamp, client, node id, response result>) is returned to the client. The client checks whether it has received at least f+1 identical results from different nodes as the final result.

[0051] The broadcast process of the master node includes three stages: pre-preparation stage, preparation stage and submission stage. The pre-preparation and preparation stages ensure that the order of request sending in the same view is correct; the preparation and submission stages ensure that the confirmation requests between different views are in order.

[0052] Pre-prepare phase: The master node assigns a proposal number to the request received from the client and then sends a pre-prepare message <<PRE-PREPARE, view, n, digest> , message> (i.e. <<pre-prepared state, view, sequence number n, message digest>, client request message>) to each replica node. The master node needs to sign the pre-prepared message. n is the sequence number assigned by the master node for this request, message is the client's request message, and digest is the digest of the message. The purpose of this step is to assign a sequence number to the client request.

[0053] Preparation phase: After receiving the pre-preparation message, the replica node checks the message (including checking the signature, view, and number). If the message is legal, it sends the preparation message to other nodes.<PREPARE, view, n, digest, id> (ie <prepare status, view, sequence number n, message digest, node id>), bring your own id information, and add a signature. The node that receives the prepare message also checks the legitimacy of the message. If the node collects at least 2f+1 verified messages, it is considered to have passed the verification and writes the prepare message to the local commit message log. This step confirms that most nodes have reached a consensus on the sequence number.

[0054] Commit phase: broadcast commit message<COMMIT, v, n, d, id> (ie <commit status, view, number n, message digest, node id>) and add your own signature to tell other nodes that a proposal numbered n has been submitted in view v. If at least 2f+1 verified commit messages are collected, it means that the proposal is accepted by the entire system.

[0055] The definitions of the terms used in the above steps are as follows: ① Replica: A node in the system. Each replica stores a copy of the system’s data and participates in the consensus process. ②Master node: The node responsible for coordinating the consensus process, receiving client requests and forwarding them to the backup node; ③ Backup nodes: All nodes except the master node, receive and verify the messages from the master node, and participate in the consensus process; ④Client: An entity that sends requests to a distributed system, expecting the system to reach consensus on its request and execute it; ⑤Preparation message: The message sent by the master node to the backup node contains the sequence number and operation content requested by the client, which is used to start the consensus process; ⑥ Prepare message: A message generated by the backup node after verifying the pre-prepare message, indicating that it has received and verified the request; ⑦ Commit message: A message generated by a node after receiving enough prepare messages, indicating that it believes that the request has reached consensus and is ready to be executed; ⑧Serial number: a unique number used to identify the client request to ensure the orderly execution of the request; ⑨ Malicious nodes: nodes that attempt to disrupt the consensus process or the normal operation of the system. PBFT allows a maximum of f malicious nodes in the system; ⑩Signature verification: When verifying a message, the node checks the digital signature of the message to ensure the authenticity of its source and the integrity of its content; In this embodiment, step S4 includes the following steps (eg Figure 6 shown): Step 4.1: Data analysis preparation: The authentication compliance analysis agent receives the integrated data packet from the cross-chain information integration agent and pre-processes the data; The preprocessing includes: structuring and formatting the data packet to ensure that the data meets the analysis requirements; checking whether the data packet contains all the necessary information, the necessary information includes the company's production records, environmental compliance documents, and supply chain traceability data; performing a preliminary check on the data's integrity and accuracy to eliminate any missing or abnormal data and ensure that the received data can provide a reliable basis for subsequent compliance analysis.

[0056] Step 4.2: Certification compliance analysis: After data preprocessing is completed, the certification compliance analysis agent uses a large language model to conduct in-depth analysis of the data to determine whether the operations and records of each company comply with the GRS (Global Recycling Standard) certification standards, and generates analysis reports and certification recommendations; The intelligent agent will analyze and understand the production records, material sources, and operation process information provided by the enterprise to evaluate whether it meets the sustainable development and environmental protection requirements of GRS. In this process, the intelligent agent will automatically identify potential compliance issues in the data and generate a detailed analysis report, which includes the enterprise's compliance score and a list of existing problems.

[0057] Step 4.3: Certification results and manual intervention: During the certification compliance analysis process, manual experts are allowed to intervene to review and verify the analysis report and certification recommendations of the certification compliance analysis agent; experts can check the accuracy and rationality of the analysis results to ensure that there are no omissions or errors in the agent's analysis process. Based on the expert's review opinions, the certification compliance analysis agent adjusts the analysis results or modifies the report content to give the final audit results; the purpose of this step is to ensure the accuracy and comprehensiveness of the audit results, and to improve the fairness and reliability of the certification process by combining the dual review mechanism of manual and intelligent.

[0058] Step 4.4: Final report generation: After confirming the final audit results, the certification compliance analysis agent generates a GRS audit report for the enterprise; the report details the enterprise’s performance during the certification process, compliance scores, and areas for improvement; if the enterprise passes all certification requirements, the certification compliance analysis agent generates a formal GRS certification certificate and records the GRS certification certificate on the regulatory blockchain to ensure the authenticity and immutability of the certificate; regulatory agencies can monitor the compliance of the enterprise in real time through the blockchain; if the enterprise fails to pass the audit, the GRS audit report is transmitted to the feedback optimization agent.

[0059] In this embodiment, step S5 includes the following steps: Figure 7 shown) Step 5.1: Receiving and analyzing audit results: The feedback optimization agent receives the GRS audit report provided by the certification compliance analysis agent; the feedback optimization agent analyzes the GRS audit report, focusing on the problems and deficiencies listed in the report, and identifies the weak links in the enterprise's certification process; the weak links include irregular behaviors in the operating procedures, incomplete records, and substandard data quality; the agent further mines and analyzes the data to understand the root causes of the company's problems in production, management or other links, and provide a basis for the formulation of subsequent improvement suggestions.

[0060] Step 5.2: Improvement suggestion generation: After analyzing the GRS audit report, the feedback optimization agent uses the large language model to generate improvement suggestions for the enterprise; the suggestions include specific operational steps and optimization strategies for specific problems of the enterprise, such as improving production processes and strengthening data management. The agent will also combine industry best practices and GRS certification standards to provide enterprises with practical improvement plans to help them improve the compliance of their processes and data quality.

[0061] Step 5.3: Suggestion transmission: After generating improvement suggestions, the feedback optimization agent transmits the improvement suggestions to the cross-chain information integration agent; the cross-chain information integration agent transmits the improvement suggestions to the relevant enterprise certification submission agent to ensure that each enterprise receives the improvement suggestions of the feedback optimization agent in a timely manner; the enterprise makes improvements and adjustments based on the improvement suggestions to improve its compliance level and data management capabilities. Through this process, the closed-loop management of feedback is ensured to promote continuous improvement and compliance optimization of enterprises.

[0062] The cross-chain authentication method for electronic waste recycling based on a multi-agent system provided by the present invention is a multi-agent system based on a large language model, and thus has advantages in complex task processing, distributed collaboration, and automated decision-making. By introducing a multi-agent system, cross-chain retrieval, integration, and verification of data can be achieved between multiple blockchains, ensuring data transparency and security during the authentication process, thereby improving the efficiency and credibility of GRS authentication.

[0063] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. The cross-chain certification system for electronic waste recycling based on a multi-agent system is characterized by: The electronic waste recycling cross-chain certification system includes: an enterprise certification submission agent, a cross-chain information integration agent, a certification compliance analysis agent, and a feedback optimization agent; The enterprise certification submission agent is used to extract data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises, process the sensitive information of the enterprises, conduct preliminary compliance checks on the processed non-sensitive data, generate an initial declaration report and submit the initial declaration report to the cross-chain information integration agent; The cross-chain information integration agent is used to perform cross-chain retrieval, integration and transmission of data between blockchains of different enterprises; The certification compliance analysis agent is used to receive the integrated data packet sent by the cross-chain information integration agent, perform compliance analysis, determine whether the corresponding enterprise meets the GRS certification standards, and provide a GRS audit report; if the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the supervision chain; if the enterprise fails to pass the audit, the GRS audit report is transmitted to the feedback optimization agent; The feedback optimization agent is used to analyze the GRS audit report provided by the certification compliance analysis agent, generate improvement suggestions, and pass the improvement suggestions to the enterprise certification submission agent through the cross-chain information integration agent to guide the enterprise to make improvements.

2. According to claim 1, the cross-chain authentication system for electronic waste recycling based on a multi-agent system is characterized in that: The cross-chain information integration agent obtains the required authentication data from the blockchains of different enterprises; the authentication data includes the initial declaration report and the supply chain data of upstream and downstream enterprises; after obtaining the data, the authentication compliance analysis agent integrates the data from different sources to ensure that all data remain synchronized; during the integration process, data consistency verification is performed to ensure that the information on each blockchain matches each other; the integrated data packet is stored on the relay blockchain after consensus and transmitted to the authentication compliance analysis agent; Furthermore, the cross-chain information integration agent has the function of feedback transmission. When the feedback optimization agent generates improvement suggestions, the cross-chain information integration agent transmits the improvement suggestion feedback to the enterprise certification submission agent.

3. According to claim 1, the cross-chain authentication system for electronic waste recycling based on a multi-agent system is characterized in that: The certification compliance analysis agent uses a large language model to conduct an in-depth analysis of the integrated data packets sent by the cross-chain information integration agent to determine whether the operations of each enterprise comply with regulatory requirements and environmental protection standards, and generate a GRS audit report; If the company passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the chain of custody; If the enterprise fails the audit, the GRS audit report is sent to the feedback optimization agent; Furthermore, the certification compliance analysis intelligent body conducts a secondary review of the data resubmitted by the enterprise after improvements.

4. An electronic waste recycling cross-chain authentication method based on a multi-agent system, using the electronic waste recycling cross-chain authentication system according to any one of claims 1 to 3, characterized in that: The electronic waste recycling cross-chain certification method comprises the following steps: S1: initialization of each agent and provision of local knowledge; S2: Data processing and initial declaration submission by the enterprise certification submission agent: The enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises, processes the sensitive information of the enterprises, performs a preliminary compliance check on the processed non-sensitive data, generates an initial declaration report and submits the initial declaration report to the cross-chain information integration agent; S3: The cross-chain information integration agent integrates the information submitted by each enterprise: the cross-chain information integration agent retrieves the required authentication data from the blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises; integrates and verifies the data on different blockchains. After obtaining the data, the authentication and compliance analysis agent integrates the data from different sources to ensure that all data remain synchronized; during the integration process, data consistency verification is performed to ensure that the information on each blockchain matches each other; the integrated data packet is stored on the relay blockchain after consensus, and then encrypted and transmitted to the authentication and compliance analysis agent; S4: The certification compliance agent generates an analysis report based on the GRS certification standards: The certification compliance analysis agent uses a large language model to conduct an in-depth analysis of the integrated data packet sent by the cross-chain information integration agent to determine whether the operations of each enterprise comply with regulatory requirements and environmental protection standards, and generate a GRS audit report. The GRS audit report shows whether the enterprise has passed the GRS certification; If the enterprise passes the audit, the certification compliance analysis agent generates the final GRS certification certificate and records it in the chain of custody to ensure that the certification results are open, transparent and cannot be tampered with, and then proceeds to step S6; If the enterprise fails the audit, the GRS audit report is transmitted to the feedback optimization agent and proceeds to step S5; S5: Feedback optimization agent analyzes audit results and proposes enterprise improvement suggestions: The feedback optimization agent receives and analyzes the GRS audit report provided by the certification compliance agent, generates improvement suggestions, and transmits the improvement suggestions to the enterprise certification submission agent through the cross-chain information integration agent to guide the enterprise to make improvements; S6: If the enterprise passes the certification, the certification process is terminated; If the enterprise certification fails, the improvement suggestions will be passed to the enterprise certification submission agent through the cross-chain information integration agent, and steps S2 to S5 will be repeated until the enterprise certification passes.

5. According to claim 4, the cross-chain authentication method for electronic waste recycling based on a multi-agent system is characterized in that: Step S1 includes: Agent initialization: allocate the required computing resources and storage space for the enterprise certification submission agent, cross-chain information integration agent, certification compliance analysis agent, and feedback optimization agent; set the communication protocol and data format between agents to ensure smooth and accurate information transmission; Local knowledge provision: Provide each agent with the necessary local knowledge base, which includes enterprise-related regulations, standard operating procedures, and historical certification data.

6. According to claim 4, the cross-chain authentication method for electronic waste recycling based on a multi-agent system is characterized in that: Step S2 includes the following steps: Step 2.1: Data retrieval and processing: The enterprise certification submission agent extracts data related to GRS certification from the respective blockchains of recycling enterprises, dismantling enterprises and remanufacturing enterprises. The data related to GRS certification includes production records and environmental compliance data; After extracting the data, the enterprise certification submission agent uses a large language model to conduct in-depth analysis of the data and identify key information in the data; Step 2.2: Sensitive information processing: The enterprise authentication submission agent analyzes and detects sensitive data, which includes information involving corporate commercial secrets or personal privacy; Enterprise authentication submission agents encrypt or specially mark sensitive data to prevent unauthorized access and leakage risks; Step 2.3: Preliminary compliance analysis: The enterprise certification submission agent conducts a preliminary compliance check on the processed non-sensitive data, compares the enterprise's operating procedures and records with the GRS certification requirements, and verifies whether the enterprise's actual operations comply with relevant standards and regulations; Step 2.4: Initial declaration submission: The enterprise certification submission agent organizes the data processed by the preliminary compliance analysis in Step 2.3 into a formatted initial declaration report, submits the initial declaration report to the cross-chain information integration agent, and waits for feedback or further processing instructions.

7. According to claim 4, the cross-chain authentication method for electronic waste recycling based on a multi-agent system is characterized in that: Step S3 includes the following steps: Step 3.1: Cross-chain data collection: The cross-chain information integration agent establishes a connection with the blockchain system of each enterprise; through the cross-chain protocol, the authentication data is collected from the blockchain of each enterprise, and the authentication data includes the initial declaration report and the supply chain data of upstream and downstream enterprises; The mathematical expression of the cross-chain data collection process is as follows: ； in, Indicates the enterprise i In time t The submitted data set, Indicates the enterprise i In time t Submitted j Class data, data from different companies' blockchains Extracted from Indicates i The blockchain of an enterprise n Indicates the number of enterprises; During the data collection process, the cross-chain information integration agent uses data verification technology to confirm the integrity and accuracy of the data, and verifies that the data has not been tampered with or lost through hash verification or digital signature means; the mathematical description is: ； in, Representation data D The hash value of Indicates the enterprise received by the cross-chain information integration agent i In time t Submitted j Class data, Representation data D Timestamp, indicating the time when the data was submitted; Step 3.2: Data integration and synchronization: During the integration process, the cross-chain information integration agent will perform format conversion and field mapping on various types of data to ensure that the format and content of all data remain consistent; and the cross-chain information integration agent will perform data synchronization checks and compare the timestamps of data submitted by each enterprise; if time differences or data conflicts are found, an automated data conflict resolution process will be performed to ensure that the final integrated data set is synchronized and conflict-free; The automated data conflict resolution process handles conflicts through a timestamp priority strategy and data verification, including the following steps: Compare the timestamps of different data and select the latest data, as shown below: ； in, For final submission of data, Representation data D timestamp; if the timestamps are the same, use hash check to confirm the integrity of the data to ensure that it has not been tampered with or lost; when data cannot be directly selected, data fusion is used to handle conflicts. The data fusion method is: Assume that the data fusion function is ; in, and are two conflicting data items, To provide a unified record without conflicts; For conflicting data of numerical type, the average value is used as the fusion data; for conflicting data of string type, the result of string concatenation is used as the fusion data, which is expressed as follows: ； Among them, Concatenate means the operation of concatenating two strings into one string; To integrate conflict-free records, use a merge operation: ； generate Final submission data, including m Indicates the total number of unified record items without conflict, i is an index, indicating the i conflict-free records, index i From 1 to m , traverse all conflict-free unified records; Step 3.3: Verification, consensus and transmission of integrated data: After completing data integration and synchronization, the final integrated data packet is verified to confirm the integrity and accuracy of all data; the data is stored in the relay blockchain after consensus using the PBFT consensus algorithm to ensure the security and traceability of the submitted data; finally, the verified and encrypted integrated data packet is transmitted to the certified compliance analysis agent.

8. According to claim 7, the cross-chain authentication method for electronic waste recycling based on a multi-agent system is characterized in that: In step 3.3, the PBFT consensus algorithm is used to achieve data consensus and then store it in the relay blockchain, including the following steps: Step 1: PBFT consensus system initialization: Set the PBFT consensus system to contain n nodes, and allow at most f nodes to be malicious nodes; require n ≥ 3f + 1, and the nodes are divided into primary nodes and backup nodes; select a node as the primary node through rotation or random algorithm, and as long as the primary node does not switch, it is called a view; Step 2: Master node broadcast: In a certain view, the client sends the request <request, operation, timestamp, client> to the master node. If the client sends it to a slave node, the slave node forwards the request to the master node. The master node is responsible for broadcasting the request to all other replica nodes and completing consensus. Step 3: Response processing: All nodes complete the request processing and return the processing result <reply, view, timestamp, client, node id, response result> to the client; when the client checks and receives at least f+1 identical results from different nodes, it takes the identical results as the final result.

9. According to claim 4, the cross-chain authentication method for electronic waste recycling based on a multi-agent system is characterized in that: Step S4 includes the following steps: Step 4.1: Data analysis preparation: The authentication compliance analysis agent receives the integrated data packet from the cross-chain information integration agent and pre-processes the data; The preprocessing includes: structuring and formatting the data packet to ensure that the data meets the analysis requirements; checking whether the data packet contains all the necessary information, including the company's production records, environmental compliance documents, and supply chain traceability data; and conducting a preliminary check on the data's completeness and accuracy; Step 4.2: Certification compliance analysis: After data preprocessing is completed, the certification compliance analysis agent uses a large language model to conduct in-depth analysis of the data, determine whether the operations and records of each enterprise comply with the GRS certification standards, and generate analysis reports and certification recommendations; Step 4.3: Certification results and human intervention: During the certification compliance analysis process, human experts are allowed to intervene to review and verify the analysis report and certification recommendations of the certification compliance analysis agent. Based on the expert's review opinion, the certification compliance analysis agent adjusts the analysis results or modifies the report content to give the final audit results. Step 4.4: Final report generation: After confirming the final audit results, the certification compliance analysis agent generates the company's GRS audit report; if the company passes all certification requirements, a formal GRS certification certificate is generated and the GRS certification certificate is recorded on the regulatory blockchain; if the company fails the audit, the GRS audit report is transmitted to the feedback optimization agent.

10. According to claim 4, the cross-chain authentication method for electronic waste recycling based on a multi-agent system is characterized in that: Step S5 includes the following steps: Step 5.1: Receiving and analyzing audit results: The feedback optimization agent receives the GRS audit report provided by the certification compliance analysis agent; the feedback optimization agent analyzes the GRS audit report and identifies the weak links in the enterprise's certification process; the weak links include irregularities in the operating process, incomplete records, and substandard data quality; Step 5.2: Improvement suggestion generation: After analyzing the GRS audit report, the feedback optimization agent uses the large language model to generate improvement suggestions for the enterprise; Step 5.3: Suggestion transmission: After generating the improvement suggestions, the feedback optimization agent transmits the improvement suggestions to the cross-chain information integration agent; The cross-chain information integration agent will pass the improvement suggestions to the relevant enterprise certification submission agent; Enterprises make improvements and adjustments based on improvement suggestions.

Images