Automatic data checking system in carbon asset transaction

Abstract

The invention provides an automatic data verification system in carbon asset transaction, and relates to the technical field of carbon asset management, the system comprises a data acquisition layer, a data integration layer, a verification model layer, an interaction and security layer, and a sliding window technology of an LSTM time sequence analysis unit increases the number of training samples. The SVM classification checking unit processes non-linear data through kernel function skills and optimizes parameters through cross validation, through the synergistic effect of the kernel function skills and the SVM classification checking unit, trend prediction and abnormal judgment of the carbon data are more reliable, misjudgment and missed judgment conditions are greatly reduced, and therefore the accuracy and reliability of the whole automatic data checking system are improved, and the data checking efficiency is improved. The method provides powerful guarantee for the fairness of the carbon asset transaction, can effectively collect and integrate the carbon asset transaction data, uses an advanced algorithm model to carry out accurate verification, guarantees the safety and credibility of data interaction through the block chain and the intelligent contract, and provides powerful technical support for the healthy and stable development of the carbon asset transaction market.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon asset management, and in particular to an automated data verification system in carbon asset trading. Background Art

[0002] According to a complex organizational structure coding system and coding method for calculating carbon asset verification disclosed in China with the publication number "CN116578289A", the organizational structure coding system includes a business management module and a collection module, a main control module is provided between the business management module and the collection module, and a data file storage unit is provided at one connection end of the main control module. In this coding system, the first digit of the code starts with C or E, C represents the organization or company level, and E represents the project level. The coding is performed according to the sorted parent-child node information. For the coding of the organization and company levels, when carbon data is calculated for carbon asset verification, the organizational structure can be quickly located according to the coding system. Then, according to the located company or project, the system can directly read the code of the project, and then find the corresponding carbon asset data, so as to realize fast and accurate search for carbon asset data of any level, any organization, and any project, with fast search speed and high accuracy.

[0003] According to a carbon asset life cycle management platform based on a new power system disclosed in China with a publication number of "CN118429135A", the platform includes a production module, a processing module, a verification module, a circulation and trading module and a post-evaluation module. Among them, the production module can collect and store the carbon emission data generated in the new power system to uniformly manage the carbon emission data from different emission sources; the processing module can process the carbon emission data into carbon assets, so that the same standard can be used to achieve the conversion of carbon assets; the verification module can verify the authenticity of the carbon asset data to ensure the authenticity and validity of the carbon assets; the circulation and trading module can connect various scene nodes and share the circulation data of carbon assets in various scene nodes, so as to ensure the transparency and legality of the circulation data; the post-evaluation module can conduct a comprehensive evaluation of the carbon emission data and various carbon assets, thereby reducing adverse factors.

[0004] The above patent documents and prior art have the following technical problems when used: Problem 1: In traditional carbon asset trading data processing, when multi-source data is integrated and model training is carried out, the original data of each data holder often needs to be brought together, which makes the data face a great risk of privacy leakage and makes it difficult to fully aggregate the data, affecting the accuracy and comprehensiveness of carbon asset trading data verification; Problem 2: Previous carbon data collection systems lacked effective pre-processing mechanisms for massive and diverse data. Large amounts of data were directly transmitted without optimization, which took up a lot of network bandwidth and had slow transmission speeds. In addition, carbon data verification models mostly used relatively simple algorithms or single algorithms, which were difficult to cope with the complex temporal and nonlinear characteristics of carbon asset trading data, and were prone to misjudgment or omission, resulting in inaccurate judgment of carbon data anomalies. Problem 3: Carbon trading data integration methods often simply pile up data together without establishing deep connections based on semantics. Data from different sources lack a unified semantic framework, making it difficult to conduct effective correlation analysis. In addition, the complex relationships between carbon trading entities and between transactions and the external environment are not explored in depth, making it impossible to fully and accurately reflect the true status of carbon assets, which is not conducive to the accurate accounting and verification of carbon asset transactions. Question 4: In the past, the functions of carbon trading smart contracts were relatively simple, tightly coupling functions such as carbon data submission, verification, and transaction settlement. When carbon trading rules or verification processes change, such as adjustments to carbon emission standards or the introduction of new trading products, the entire smart contract code needs to be modified on a large scale. This not only has high development costs, but also easily introduces new errors, making it difficult for smart contracts to adapt to the dynamic changes in the carbon trading market. At the same time, the execution process of trading smart contracts lacks a transparent supervision mechanism, making it difficult for external regulatory agencies and third-party audits to gain an in-depth understanding of the details of contract execution, and unable to promptly discover loopholes or violations in contract execution. Summary of the invention

[0005] Technical issues solved In view of the shortcomings of the prior art, the present invention provides an automated data verification system in carbon asset trading, which solves the following problems: 1. Regarding the problem of insufficient data privacy protection in the carbon trading process; 2. Address the problem of poor efficiency and accuracy in carbon trading data processing; 3. The lack of deep semantic association in carbon trading data integration and the difficulty in mining hidden information in complex transaction relationships; 4. Address the issues of single functionality and difficult maintenance of smart contracts in the carbon trading process.

[0006] Technical Solution

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an automated data verification system in carbon asset trading, the system includes a data collection layer, a data integration layer, a verification model layer, and an interaction and security layer, wherein: The data acquisition layer is used to obtain carbon asset trading related data from multiple sources and perform preliminary processing, including a federated learning and edge computing module, which is used to realize the collaborative training and verification model of multiple data holders through a federated learning algorithm and pre-process data at the data source end using edge computing. The federated learning and edge computing module includes a federated learning subunit and an edge computing subunit. The federated learning subunit is used to establish a federated learning coordination server to communicate and coordinate with the local nodes of each data provider. Each data provider trains a local model based on local carbon data and encrypts and uploads the model parameters to the coordination server. After the server aggregates the parameters, it sends down an updated local model to iteratively train the global model. The edge computing subunit is used to deploy edge computing equipment near the internal data source of the enterprise to perform real-time pre-processing on the raw data, including data compression, format conversion and preliminary screening. The screened suspicious abnormal data is transmitted to the data integration layer; The data integration layer is used to integrate, associate, store and manage the collected data based on semantic web technology, including a semantic web data integration module, which includes an ontology construction unit, a semantic mapping and association unit, and a data storage and management unit. The ontology construction unit is used to construct an ontology model in the field of carbon trading to define core concepts and their interrelationships. The semantic mapping and association unit is used to map and match the collected data with the ontology model and associate and integrate them. The data storage and management unit is used to classify and store the integrated data and manage access rights. The verification model layer is used to verify the integrated data using multiple algorithm models, including a fused LSTM-SVM verification module and a GNN relationship verification module. The fused LSTM-SVM verification module includes an LSTM time series analysis unit and an SVM classification verification unit. The LSTM time series analysis unit is used to obtain historical carbon data and related time series data to train the LSTM network model to output a time series feature vector and a predicted trend value. The SVM classification verification unit is used to receive the feature vector and determine whether the data is normal based on the trained SVM model. The GNN relationship verification module includes a carbon trading relationship graph construction unit and a GNN analysis and verification unit. The carbon trading relationship graph construction unit is used to construct a carbon trading relationship graph. The GNN analysis and verification unit is used to mine abnormal patterns based on the relationship graph using the GNN algorithm; The interaction and security layer is used to ensure the security and reliability of data interaction and automatically execute transactions and verification processes, including a blockchain and smart contract module. The blockchain and smart contract module includes a blockchain data storage and interaction unit and a smart contract execution unit. The blockchain data storage and interaction unit is used to establish a blockchain network to store key data and provide a query interface. The smart contract execution unit is used to write smart contract code and automatically execute carbon asset trading and verification related operations based on conditions.

[0008] Preferably, in the federated learning subunit, each data provider's local node uses homomorphic encryption technology to encrypt local model parameters to ensure that data privacy is not leaked during parameter upload, and the coordination server uses a multi-party computing protocol to aggregate the encrypted parameters to achieve secure global model parameter updates.

[0009] Preferably, during the data compression process, the edge computing subunit adopts an adaptive compression algorithm according to the time series characteristics and data type of the data, using a higher compression ratio for basic data that changes slowly and a lower compression ratio for real-time monitoring data that fluctuates greatly, thereby minimizing the amount of data transmission while ensuring data validity.

[0010] Preferably, when constructing the ontology model in the field of carbon trading, the ontology construction unit adopts a construction method that combines top-down and bottom-up approaches based on internationally accepted carbon trading standards and industry specifications, first determining the core concept framework, then gradually refining the sub-concepts and attributes, and regularly updating and maintaining the ontology model according to changes in policies and regulations and industry development trends.

[0011] Preferably, the semantic mapping and association unit adopts a mapping method that combines rule-based and machine learning-based mapping when performing semantic mapping. Predefined rules are used to quickly map common data formats and standard terms, and machine learning models are used to automatically learn and map fuzzy data expressions.

[0012] Preferably, when training the LSTM network model, the LSTM time series analysis unit uses a sliding window technique to segment the historical carbon data to increase the number of training samples, and combines the early stopping method to prevent overfitting, while introducing an attention mechanism so that the model can focus on the time series feature parts that have an important impact on carbon verification.

[0013] Preferably, when constructing the classification hyperplane, the SVM classification verification unit uses the kernel function technique to map the nonlinearly separable data, selects the radial basis function as the kernel function, and optimizes the kernel function parameters and penalty factors through the cross-validation method to improve the accuracy of the SVM model in classifying complex carbon data.

[0014] Preferably, when constructing the carbon trading relationship graph, the carbon trading relationship graph construction unit adds basic trading entities, trading relationships, carbon trading-related policy and regulatory changes, and market dynamic information into the relationship graph as virtual nodes and edges to more comprehensively reflect the impact of the carbon trading environment on trading relationships.

[0015] Preferably, the GNN analysis and verification unit adopts a multi-layer GNN structure when performing information propagation and node feature update. Each layer of GNN adopts different message passing functions and aggregation functions, and dynamically adjusts the scope and intensity of information propagation according to the node type and edge weight, so as to more accurately mine abnormal patterns in complex relationships.

[0016] Preferably, the smart contract execution unit adopts a modular design when writing the smart contract code, and encapsulates functions such as carbon data submission, verification process, transaction settlement, and violation penalties into independent modules to facilitate contract maintenance and upgrades. At the same time, an audit mechanism for smart contracts is set up to allow regulatory agencies and third-party audit agencies to supervise and audit the contract execution process.

[0017] Beneficial Effects

[0018] The present invention provides an automated data verification system for carbon asset trading, which has the following beneficial effects: 1. The system of the present invention adopts the homomorphic encryption technology of the local node of the data provider in the federated learning subunit and the multi-party computing protocol of the coordination server. In the entire data collection and model training process, the data privacy of each participant is strongly protected. Enterprises and other data holders do not need to worry about the leakage of sensitive carbon data information, and can safely participate in the construction of the global verification model, ensuring that the confidentiality and integrity of the data are still maintained even in the complex environment of multi-source data collaborative training. It meets the extremely high data security requirements in the field of carbon trading, effectively avoids the risks of exposure of commercial secrets and malicious tampering of carbon data due to data leakage, and provides solid support for the stability and trust foundation of the carbon asset trading market.

[0019] 2. The system of the present invention uses the edge computing subunit of the data acquisition layer to adopt an adaptive compression algorithm, and flexibly adjusts the compression ratio according to the time series characteristics and type of the data. For a large amount of basic data that changes slowly, a high compression ratio reduces the transmission bandwidth demand and storage pressure, while for real-time monitoring data with large fluctuations, a lower compression ratio ensures that the key details of the data are retained. This makes data transmission more efficient and the data quality is not damaged, providing a timely and accurate data basis for the subsequent verification process, accelerating the data flow speed of the entire system, and the sliding window technology of the LSTM time series analysis unit increases the number of training samples, combined with the early stopping method to prevent overfitting and the attention mechanism to focus on key features, making the learning of the time series characteristics of carbon data more accurate. The SVM classification verification unit processes nonlinear data through kernel function techniques and optimizes parameters through cross-validation, effectively improving the accuracy of complex carbon data classification. The two work together, whether it is the trend prediction of carbon data or the judgment of anomalies, it is more reliable, greatly reducing misjudgments and missed judgments, thereby improving the accuracy and reliability of the entire automated data verification system, and providing a strong guarantee for the fairness and justice of carbon asset transactions.

[0020] 3. The system of the present invention adopts the ontology construction unit of the data integration layer to construct the ontology model of the carbon trading field by combining the top-down and bottom-up methods, and updates and maintains it according to policies, regulations and industry development dynamics. This enables the system to effectively integrate the collected multi-source data based on a comprehensive and up-to-date conceptual framework, clarify the internal connection between each data, break the data island phenomenon, and achieve deep integration at the data semantic level, laying a good foundation for accurate carbon asset accounting and transaction verification. The GNN relationship verification module incorporates a variety of key elements when constructing the carbon trading relationship diagram, and analyzes it through a multi-layer GNN structure and a dynamically adjusted information dissemination mechanism. It can not only dig out the potential associations and abnormal patterns between trading entities, but also comprehensively consider the impact of external environmental factors on carbon trading relationships, which helps regulators to grasp the overall health of the carbon trading market from a macro level, timely discover systemic risks and illegal trading behaviors, and promote the standardization and sustainable development of the carbon trading market.

[0021] 4. The system of the present invention utilizes a modular design of the smart contract execution unit to independently encapsulate the core functions related to carbon data into modules. This design makes the structure of the contract clearer and enhances the independence of the functional modules. When the carbon trading rules or verification process changes, only the corresponding modules need to be modified or upgraded in a targeted manner without making large-scale changes to the entire smart contract, which greatly reduces the complexity and cost of contract maintenance. At the same time, the audit mechanism set up allows regulators and third-party auditing agencies to intervene to supervise and audit the contract execution process, further enhancing the transparency and credibility of smart contracts, ensuring that the carbon asset trading and verification process strictly follows the preset rules, effectively preventing loopholes in the contract execution process from being exploited, improving the security and credibility of the entire carbon trading system, and promoting the healthy and orderly operation of the carbon asset trading market. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a system architecture diagram of the present invention; Figure 2 It is the system data trend diagram of the present invention; Figure 3 This is a diagram of the steps of the system carbon asset trading of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific embodiment 1

[0025] like Figure 1-3 As shown, an automated data verification system in carbon asset trading includes a data collection layer, a data integration layer, a verification model layer, and an interaction and security layer, wherein: The data collection layer is used to obtain carbon asset trading related data from multiple sources and perform preliminary processing, including a federated learning and edge computing module, which is used to implement collaborative training and verification models by multiple data holders through a federated learning algorithm and pre-process data at the data source end using edge computing. The federated learning and edge computing module includes a federated learning subunit and an edge computing subunit; The data integration layer is used to integrate, associate, store and manage the collected data based on semantic web technology, including a semantic web data integration module, which includes an ontology construction unit, a semantic mapping and association unit, and a data storage and management unit; The verification model layer is used to verify the integrated data using a variety of algorithm models, including the fused LSTM-SVM verification module and the GNN relationship verification module. The fused LSTM-SVM verification module includes an LSTM time series analysis unit and an SVM classification verification unit. The LSTM time series analysis unit is used to obtain historical carbon data and related time series data to train the LSTM network model to output time series feature vectors and predicted trend values. The SVM classification verification unit is used to receive feature vectors and determine whether the data is normal based on the trained SVM model. The GNN relationship verification module includes a carbon trading relationship graph construction unit and a GNN analysis and verification unit. The carbon trading relationship graph construction unit is used to construct a carbon trading relationship graph. The GNN analysis and verification unit is used to mine abnormal patterns based on the relationship graph using the GNN algorithm. The interaction and security layer is used to ensure the security and reliability of data interaction and to automate the transaction and verification processes, including the blockchain and smart contract modules. The blockchain and smart contract modules include a blockchain data storage and interaction unit and a smart contract execution unit. The blockchain data storage and interaction unit is used to establish a blockchain network to store key data and provide a query interface. The smart contract execution unit is used to write smart contract code and automatically execute carbon asset trading and verification related operations based on conditions.

[0026] This automated data verification system can effectively collect and integrate carbon asset trading data, use advanced algorithm models for accurate verification, and ensure the security and credibility of data interaction through blockchain and smart contracts, providing strong technical support for the healthy and stable development of the carbon asset trading market. Specific embodiment 2

[0028] like Figure 1-3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed: For each unit module in each level of the system, the following contents are further included: The federated learning subunit is used to establish a federated learning coordination server to communicate and coordinate with the local nodes of each data provider. Each data provider trains a local model based on local carbon data and encrypts the model parameters and uploads them to the coordination server. The server aggregates the parameters and sends the updated local model to iteratively train the global model. In the federated learning subunit, the local nodes of each data provider use homomorphic encryption technology to encrypt the local model parameters to ensure that data privacy is not leaked during the parameter upload process, and the coordination server uses a multi-party computing protocol to aggregate the encrypted parameters to achieve secure global model parameter updates. Encryption technology and multi-party computing protocols that coordinate servers ensure that the data privacy of all participants is strongly protected throughout the data collection and model training process. Enterprises and other data holders do not need to worry about the leakage of sensitive carbon data information and can safely participate in the construction of the global verification model. This encryption and security aggregation mechanism ensures that even in the complex environment of multi-source data collaborative training, the confidentiality and integrity of the data are maintained, which meets the extremely high data security requirements in the carbon trading field, and effectively avoids the risks of exposure of commercial secrets and malicious tampering of carbon data due to data leakage, providing solid support for the stability and trust foundation of the carbon asset trading market.

[0029] The edge computing subunit is used to deploy edge computing devices near the internal data source of the enterprise to perform real-time pre-processing of the original data, including data compression, format conversion and preliminary screening. The screened suspicious abnormal data is transmitted to the data integration layer. During the data compression process, the edge computing subunit adopts an adaptive compression algorithm according to the time series characteristics and data type of the data. A higher compression ratio is used for basic data that changes slowly, and a lower compression ratio is used for real-time monitoring data with large fluctuations. While ensuring the validity of the data, the data transmission volume is minimized to the greatest extent. The edge computing subunit adopts an adaptive compression algorithm to flexibly adjust the compression ratio according to the time series characteristics and type of the data. For a large amount of basic data that changes slowly, a high compression ratio reduces the transmission bandwidth demand and storage pressure; and for real-time monitoring data with large fluctuations, a lower compression ratio ensures that the key details of the data are retained, which makes data transmission more efficient and the data quality is not damaged, providing a timely and accurate data basis for the subsequent verification process and accelerating the data flow speed of the entire system.

[0030] The ontology construction unit is used to construct an ontology model in the field of carbon trading to define core concepts and their relationships. When constructing an ontology model in the field of carbon trading, the ontology construction unit adopts a top-down and bottom-up construction method based on internationally accepted carbon trading standards and industry specifications. It first determines the core concept framework, then gradually refines the sub-concepts and attributes, and regularly updates and maintains the ontology model according to changes in policies and regulations and industry development trends. The semantic mapping and association unit is used to map and match the collected data with the ontology model and associate and integrate them. When performing semantic mapping, the semantic mapping and association unit adopts a mapping method that combines rule-based and machine learning-based mapping. For common data formats and Standard terms are quickly mapped using predefined rules, and fuzzy data expressions are automatically learned and mapped using machine learning models. The data storage and management unit is used to classify and store integrated data and manage access rights. The ontology construction unit uses a combination of top-down and bottom-up methods to construct an ontology model in the carbon trading field, and updates and maintains it based on policies, regulations and industry development dynamics. This enables the system to effectively integrate the collected multi-source data based on a comprehensive and up-to-date conceptual framework, clarify the inherent connections between the data, break the data silos, and achieve deep integration at the data semantic level, laying a good foundation for accurate carbon asset accounting and transaction verification.

[0031] When training the LSTM network model, the LSTM time series analysis unit uses the sliding window technology to segment the historical carbon data to increase the number of training samples, and combines the early stopping method to prevent overfitting. At the same time, the attention mechanism is introduced to enable the model to focus on the time series feature parts that have an important impact on carbon verification. When constructing the classification hyperplane, the SVM classification verification unit uses the kernel function technique to map the nonlinearly separable data, selects the radial basis function as the kernel function, and optimizes the kernel function parameters and penalty factors through the cross-validation method to improve the accuracy of the SVM model in classifying complex carbon data. The sliding window technology of the LSTM time series analysis unit increases the number of training samples, combines the early stopping method to prevent overfitting and the attention mechanism to focus on key features, making the learning of the time series features of carbon data more accurate. The SVM classification verification unit processes nonlinear data through the kernel function technique and optimizes the parameters through cross-validation, which effectively improves the accuracy of complex carbon data classification. The synergistic effect of the two makes both the trend prediction and abnormal judgment of carbon data more reliable, greatly reducing the misjudgment and missed judgment, thereby improving the accuracy and reliability of the entire automated data verification system, and providing a strong guarantee for the fairness and justice of carbon asset trading.

[0032] When constructing the carbon trading relationship diagram, the carbon trading relationship diagram construction unit adds the basic trading subjects, trading relationships, changes in carbon trading policies and regulations, and market dynamic information as virtual nodes and edges into the relationship diagram to more comprehensively reflect the impact of the carbon trading environment on trading relationships. The GNN analysis and verification unit adopts a multi-layer GNN structure when performing information dissemination and node feature update. Each layer of GNN adopts different message passing functions and aggregation functions, and dynamically adjusts the scope and intensity of information dissemination according to the node type and edge weight to more accurately mine abnormal patterns in complex relationships. The GNN relationship verification module incorporates a variety of key elements when constructing the carbon trading relationship diagram, including trading subjects, relationships, changes in policies and regulations, and market dynamic information, and analyzes them through a multi-layer GNN structure and a dynamically adjusted information dissemination mechanism. This can not only mine the potential associations and abnormal patterns between trading subjects, but also comprehensively consider the impact of external environmental factors on carbon trading relationships, which helps regulators grasp the overall health of the carbon trading market from a macro level, timely discover systemic risks and illegal trading behaviors, and promote the standardization and sustainable development of the carbon trading market.

[0033] When writing smart contract codes, the smart contract execution unit adopts modular design, encapsulating functions such as carbon data submission, verification process, transaction settlement, and violation penalties into independent modules, which is convenient for contract maintenance and upgrade. At the same time, an audit mechanism for smart contracts is set up to allow regulators and third-party audit institutions to supervise and audit the contract execution process. The smart contract execution unit adopts modular design, and independently encapsulates the core functions related to carbon data into modules. This design method makes the structure of the contract clearer and the independence of the functional modules is enhanced. When the carbon trading rules or verification process changes, only the corresponding modules need to be modified or upgraded in a targeted manner without making large-scale changes to the entire smart contract, which greatly reduces the complexity and cost of contract maintenance. At the same time, the audit mechanism set up allows regulators and third-party audit institutions to intervene and supervise and audit the contract execution process, further enhancing the transparency and credibility of smart contracts, ensuring that the carbon asset trading and verification process strictly follows the preset rules, effectively preventing loopholes in the contract execution process from being exploited, improving the security and credibility of the entire carbon trading system, and promoting the healthy and orderly operation of the carbon asset trading market. Specific embodiment three: like Figure 1-3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed: In the whole system, the algorithms and model contents at each level further include the following: The formula of the data collection layer-federated learning and edge computing module is as follows: Federated learning subunit - homomorphic encryption related, encryption process principle: suppose the local model parameters to be encrypted by the local node of the data provider are , the homomorphic encryption function is , the encrypted parameters are expressed as Here, the characteristic of homomorphic encryption is that for certain specific operations, such as addition and multiplication, which involve model parameter aggregation related operations in this scenario, the operation can be performed directly on the encrypted data, and the result obtained after decryption is the same as the result of the same operation on the plaintext data. For example, if there are two encrypted parameters and , and perform addition operation to obtain The addition here is the definition of addition under the homomorphic encryption operation system. The decrypted result should be equal to the plaintext result.

[0035] Implementation steps: After training the local model, the local node of the data provider obtains the local model parameters , using homomorphic encryption function right Encrypt and get , the encrypted parameters Upload to the coordination server.

[0036] Operation logic: Each data provider independently trains the local model, generates local model parameters, and then encrypts the parameters through homomorphic encryption and uploads them. After receiving the encrypted parameters, the coordination server uses the multi-party computing protocol to perform aggregation operations. During the whole process, the data is always transmitted and processed in encrypted form until the local node needs to update the local model according to the update sent by the coordination server. The decryption operation is not emphasized here and can be understood as the inverse process of homomorphic encryption.

[0037] Beneficial effects: It ensures that data privacy is not leaked during the parameter upload process, so that all data providers can safely participate in the construction of the global verification model. Even in the complex environment of multi-source data collaborative training, the confidentiality and integrity of the data can be maintained, effectively avoiding the risks of exposure of commercial secrets and malicious tampering of carbon data caused by data leakage.

[0038] The edge computing subunit - adaptive compression algorithm includes the following: Operation logic: Each data provider independently trains the local model, generates local model parameters, and then encrypts the parameters through homomorphic encryption and uploads them. After receiving the encrypted parameters, the coordination server uses the multi-party computing protocol to perform aggregation operations. During the whole process, the data is always transmitted and processed in encrypted form until the local node needs to update the local model according to the update sent by the coordination server. The decryption operation is not emphasized here and can be understood as the inverse process of homomorphic encryption.

[0039] Beneficial effects: It ensures that data privacy is not leaked during the parameter upload process, so that all data providers can safely participate in the construction of the global verification model. Even in the complex environment of multi-source data collaborative training, the confidentiality and integrity of the data can be maintained, effectively avoiding the risks of exposure of commercial secrets and malicious tampering of carbon data caused by data leakage.

[0040] The edge computing subunit - adaptive compression algorithm includes the following: Assume the original data volume is , the compression ratio is , the amount of data after compression is ,but For basic data that changes slowly, let the corresponding compression ratio be , a higher value, such as , which means compression to 20% of the original value. For real-time monitoring data with large fluctuations, the corresponding compression ratio is , the value is lower, such as , which means compression to 50% of the original; in: : represents the size of the original data, that is, the amount of information of the uncompressed original carbon data obtained from the internal data source of the enterprise, and the unit can be bytes, etc.; : Compression ratio, which indicates the degree to which the original data is compressed. The value range is between and . The closer it is, the higher the compression degree is and the smaller the amount of compressed data is. : The amount of data after compression is obtained by multiplying the original data amount by the compression ratio; and : Compression ratios are set for different types of data, such as slowly changing basic data and real-time monitoring data with large fluctuations. Different compression ratios are selected according to the characteristics of the data to minimize the amount of data transmission while ensuring data validity.

[0041] Implementation steps: First, determine the type of data, whether it is basic data that changes slowly or real-time monitoring data with large fluctuations. If it is basic data that changes slowly, determine the compression ratio. , then according to the formula Calculate the amount of data after compression. If it is real-time monitoring data with large fluctuations, determine the compression ratio , then according to the formula Calculate the amount of compressed data and transfer the compressed data to the data integration layer.

[0042] Operation logic: The edge computing devices deployed near the enterprise's internal data sources will determine the type of the acquired raw data in real time, select the corresponding compression ratio according to different types, and then calculate the compressed data volume according to the compression formula and transmit it. In this way, the compression degree can be flexibly adjusted according to the actual situation of the data, which not only reduces the transmission bandwidth demand and storage pressure (for basic data that changes slowly), but also ensures that the key details of the data are retained (for real-time monitoring data with large fluctuations).

[0043] Beneficial effects: It makes data transmission more efficient and does not affect data quality, provides a timely and accurate data basis for subsequent verification processes, and accelerates the data flow speed of the entire system.

[0044] Data Integration Layer - Algorithms and models in the semantic web data integration module further include the following: Ontology construction unit - ontology model construction: Assume that the core concept set in the field of carbon trading is , the relationship set between each core concept is , here the relationship From the concept To another concept , then the carbon trading domain ontology model ; in: : Represents the core concept set in the field of carbon trading. These core concepts are abstract summaries of carbon trading related things and phenomena, such as carbon emissions and carbon quotas of carbon trading entities, and are the basic elements for building ontology models; : Represents the set of relationships between core concepts. It describes how different core concepts are related to each other. For example, there may be a possession relationship between carbon trading entities and carbon emissions. The relationship here is expressed in a similar form, indicating a certain connection from one concept to another.

[0045] :It is the ontology model of carbon trading, which is composed of core concepts and relationship set The overall structure provides a clear framework for subsequent data integration and semantic association by clarifying these core concepts and their interrelationships.

[0046] Implementation steps: Based on the internationally accepted carbon trading standards and industry norms, first determine the core concept set in the field of carbon trading , extract key conceptual elements through analysis of relevant standards and specifications, and then determine the relationship set between each core concept , through the analysis of actual carbon trading business processes and scenarios, find out the various connections between different concepts, and finally build a carbon trading domain ontology model , regularly update and maintain the ontology model according to changes in policies and regulations and industry development trends, that is, review and adjust the core concept set and relationship set.

[0047] Operation logic: First, obtain information from established standards and specifications as well as actual business scenarios to determine core concepts and the relationships between them, and then combine these elements into an ontology model. Over time, due to changes in policies, regulations and industry development, this ontology model needs to be dynamically updated to ensure that it can accurately reflect the current situation in the carbon trading field, thereby providing an effective framework for subsequent operations such as data integration.

[0048] Beneficial effects: The system can effectively integrate the collected multi-source data based on a comprehensive and up-to-date conceptual framework, clarify the internal connections between the data, break the data silos, and achieve deep integration of data semantics, laying a good foundation for accurate carbon asset accounting and transaction verification.

[0049] Semantic mapping and association units - mapping methods based on a combination of rules and machine learning (non-typical mathematical formulas, mainly based on method descriptions); Assume that the collected data set is , the ontology model is (defined previously), the mapping result set is For common data formats and standard terms, predefined rules are used for mapping. Let the predefined rule function be , then for this type of data, the mapping result , fuzzy data representation, assuming the machine learning model is , then the mapping result ; in: : Represents the carbon trading related data set collected from the data collection layer. These data may have different formats and expressions and need to be mapped and matched with the ontology model; : The ontology model of the carbon trading field is the model built previously to provide a conceptual framework and relationship description, which serves as the target framework for mapping; : Mapping result set, which records the collection data set Each data element in the ontology model The mapped results are obtained through different mapping methods (rule-based or machine learning-based); : Predefined rule functions, for common data formats and standard terms, map the data according to pre-set rules, match it with the corresponding elements in the ontology model, and obtain the mapping results ; :Machine learning model, fuzzy data representation, using machine learning model to analyze data Perform automatic learning and mapping matching to obtain mapping results .

[0050] Implementation steps: For the collection of data sets Each data element in First, determine whether it belongs to a common data format and standard terminology. If so, use the predefined rule function Map and get the mapping result , and add it to the mapping result set If not, use a machine learning model Map and get the mapping result , and add it to the mapping result set Finally, we get a complete set of mapping results. , complete the mapping, matching and association integration between the collected data and the ontology model.

[0051] Operation logic: For each collected data element, according to its data format and expression characteristics, select the appropriate mapping method (rule-based or machine learning-based) for mapping operation, and match it with the corresponding element in the ontology model, so as to achieve effective mapping and associative integration of data and ontology model, so as to facilitate subsequent data storage and management operations.

[0052] Beneficial effects: The collected data can be accurately mapped, matched and associated with the ontology model, and good mapping effects can be obtained for common data formats and standard terms, as well as new or ambiguous data expressions, thereby providing an accurate basis for subsequent data storage and management, and further promoting deep integration of data semantics.

[0053] Verification model layer - fusion LSTM-SVM verification module: LSTM time series analysis unit-LSTM network model training further includes the following: The core formula of LSTM mainly involves several gating mechanisms inside it, including the input gate , Forget Gate , output gate and cell status ; Forget Gate: ; in, is the sigmoid function, is the weight matrix of the forget gate, is the hidden state at the previous moment, is the input data at the current moment, is the bias vector of the forget gate; Input Gate: ; Among them, the meaning of each parameter is similar to that of the forget gate. is the weight matrix of the input gate is the bias vector of the input gate; Cell status update: ; in, is the hyperbolic tangent function, is the weight matrix used to update the cell state, is its bias vector; Output Gate: ; Among them, the meaning of each parameter is similar to that of the forget gate. is the weight matrix of the output gate, is the bias vector of the output gate; Final hidden state: ; in: : The sigmoid function, whose formula is , which maps the input value to the interval , used to control the flow and selection of information, and plays a key role in various gating mechanisms of LSTM, such as deciding which information should be forgotten (forget gate) and which information should be input (input gate), etc.; , , , : They are the forget gate, input gate, weight matrix for updating cell state and output gate. These weight matrices determine the influence of input data and the hidden state at the previous moment on each gating mechanism and cell state update. The values ​​of these weight matrices are continuously adjusted during the training process so that the LSTM model can better learn the characteristics of the data. , , , : They are the forget gate, input gate, bias vector for updating cell state and output gate. Together with the weight matrix, they affect the calculation results of each gating mechanism, and their values ​​are also adjusted continuously during the training process; : The hidden state at the previous moment, which reflects the model's learning and processing results of the data at the previous moment, and is an important reference for calculating the various gating mechanisms and cell state updates at the current moment; : The input data at the current moment, that is, the data points at the current moment obtained from the historical carbon data and related time series data, is the object of learning and processing by the LSTM model; : Cell state, which is an important state variable in the LSTM model. It records some key information of the model at different times and is continuously updated through mechanisms such as forget gate and input gate to adapt to data changes and learn the characteristics of data; : The final hidden state, which is the output result of the model processing the data at the current moment, and is also the reference for calculating the update of each gating mechanism and cell state at the next moment. By continuously iteratively calculating the hidden state at each moment, the LSTM model can learn the time series characteristics of the data.

[0054] Implementation steps: First, obtain historical carbon data and related time series data as input data , and initialize the weight matrix of each gating mechanism , , , , and the bias vector , , , , and the initial hidden state , every moment , calculate the forget gate in turn according to the above formula , Input Gate , cell state update , output gate and the final hidden state In the calculation process, the historical carbon data is segmented by sliding window technology to increase the number of training samples. For example, assuming that the original historical carbon data sequence is , the sliding window size is , multiple training samples can be obtained, such as Etc., combined with the early stopping method to prevent overfitting, that is, during the training process, when the loss function on the validation set no longer decreases or decreases very little (below a set threshold), stop training.

[0055] The attention mechanism is introduced during the training process. The calculation of the attention mechanism is as follows: Let the hidden state sequence be ( is the sequence length), the attention weight is calculated as

[0056] in, , , It is a trainable parameter. Through the attention mechanism, the model can focus on the time series features that have an important impact on carbon verification, and obtain the weighted hidden state representation:

[0057] The LSTM network model is trained until the training stop condition is met (such as reaching the set number of training rounds or the early stopping method is triggered), and a trained LSTM model is obtained, which can output time series feature vectors and predicted trend values.

[0058] Operation logic: At each time step, LSTM and the hidden state at the previous moment , through the gating mechanism to control the cell state Update and calculate the hidden state at the current moment The sliding window technology enables the model to use more historical data information for learning, increasing the diversity and quantity of training samples. The attention mechanism calculates the attention weight , the hidden states at different times Weighted summation is performed to highlight the feature information that is important for carbon verification, enabling the model to better capture key patterns and trends in the data. Finally, the trained LSTM model can output vectors reflecting the time series characteristics of the data and predicted values ​​for future carbon data trends based on the input historical carbon data sequence, providing valuable input for subsequent SVM classification verification.

[0059] Beneficial effects: The sliding window technology of the LSTM time series analysis unit increases the number of training samples. Combined with the early stopping method to prevent overfitting and the attention mechanism to focus on key features, the learning of the time series characteristics of carbon data is more accurate. This helps to more accurately grasp the changing patterns and trends of carbon data in the time dimension, and provides a reliable basis for subsequent judgment of whether the data is abnormal, which improves the accuracy and reliability of the entire automated data verification system in analyzing the time series characteristics of carbon data.

[0060] SVM classification verification unit - SVM model construction and classification content is as follows: For the linearly separable case, the decision function of SVM is

[0061] in, is the weight vector, is the input feature vector (here, the time series feature vector output by LSTM), is the bias term; The goal is to find the optimal and , so that the constraints are met, Minimum, the constraint is that for the training sample , is the category label of the sample, such as normal or abnormal:

[0062] For the case of nonlinear separability, the kernel function technique is used to map the input data to a high-dimensional space. Let the kernel function be

[0063] in, is the mapping function; The commonly used radial basis function RBF kernel function is

[0064] in, is the kernel parameter, then the decision function becomes

[0065] in, is the Lagrange multiplier obtained by solving the quadratic programming problem; explain: : Weight vector, which determines the direction of the decision boundary in the feature space. Its value is adjusted through the training process so that the decision boundary can correctly separate samples of different categories. In the case of nonlinear separability, although the data cannot be linearly divided in the original space, after mapping the data to a high-dimensional space through the kernel function, It plays a similar role of partitioning in high-dimensional space; : Input feature vector, which is the time series feature vector output by the LSTM time series analysis unit, containing the time series feature information of carbon data. SVM determines the category of the data based on these feature vectors; : Bias term, which together with the weight vector determines the position of the decision boundary in the feature space, has an important influence on the classification results, and will also be optimized and adjusted during the training process; : The category label of the sample, which is used to indicate whether the carbon data sample is normal or abnormal, is the supervisory information for SVM model training. The model parameters are optimized by making the classification results match these labels as much as possible.

[0066] : Kernel function, for nonlinearly separable data, it maps the data points in the original space to the high-dimensional feature space, so that the data in the high-dimensional space may become linearly separable. Different kernel functions have different mapping methods. The RBF kernel function is a commonly used kernel function. The greater the complexity of the feature space after mapping through parameter control, the stronger the model's ability to fit the training data, but it may also lead to overfitting, which needs to be optimized through methods such as cross-validation; : Lagrange multiplier, a variable introduced when solving the SVM optimization problem (whether linear or nonlinear), its value is determined by solving the quadratic programming problem, and then the value of the weight vector and bias term is determined.

[0067] Implementation steps: Receive the feature vector output by the LSTM time series analysis unit For linearly separable data, the weight vector is solved directly according to the optimization objective and constraints of the linear SVM. and the bias term , build a decision function sign For nonlinearly separable data, radial basis function is selected as the kernel function, and the kernel function parameters are optimized by cross-validation method. and penalty factors (related to solving the constraints in the quadratic programming problem). The specific approach is to divide the training data into a training set and a validation set. Train the SVM model under the combination of the penalty factor value, evaluate the performance of the model (such as accuracy, recall, etc.) on the validation set, select the best performance parameter combination, and then build the decision function , classify the input feature vector according to the constructed decision function to determine whether the data is normal.

[0068] Operation logic: The SVM classification verification unit takes the feature vector output by LSTM as input, and first determines whether the data is linearly separable. If so, the optimal weight vector and bias term are solved according to the principle of linear SVM, and a decision function is constructed for classification. If the data is nonlinearly separable, the kernel function is used to map the data to a high-dimensional space, and the kernel function parameters and penalty factors are optimized through cross-validation, and a decision function based on the kernel function is constructed for classification. In the whole process, the SVM model divides the feature vector into normal and abnormal categories by finding the optimal decision boundary, thereby realizing the judgment of whether the carbon data is normal.

[0069] Beneficial effects: The SVM classification verification unit processes nonlinear data through kernel function techniques and optimizes parameters through cross-validation, which effectively improves the accuracy of complex carbon data classification. It can accurately judge whether there are anomalies in carbon data based on the time series feature vector extracted by LSTM, and works together with the LSTM time series analysis unit. It is more reliable for both trend prediction and anomaly judgment of carbon data, greatly reducing misjudgments and missed judgments, further improving the accuracy and reliability of the entire automated data verification system, and providing a strong guarantee for the fairness and justice of carbon asset transactions.

[0070] Verification model layer-GNN relationship verification module Carbon trading relationship diagram construction unit - Carbon trading relationship diagram construction: Assume that the carbon trading relationship diagram is , Among them, the transaction subject set is ; The transaction relationship set is ; Information collection on policy and regulatory changes and market dynamics is , then the node set ; Edge Set It includes the transaction relationship edges between transaction entities and the association edges between transaction entities and policy and regulatory changes and market dynamic information. For example, if a transaction entity Affected by changes in policies and regulations If the influence of ; in: : Carbon trading relationship diagram, which is a graph structure used to represent the relationship between various elements in carbon trading, and describes the complex environment of carbon trading through nodes and edges; : The trading subject set includes all the subjects participating in carbon trading, such as enterprises, emission trading institutions, etc. These subjects are the core participants of carbon trading and are important node types for constructing relationship graphs; : A set of transaction relationships, which describes various transaction relationships between transaction entities, such as the buying and selling relationship of carbon quotas, the transfer relationship of carbon credits, etc. These relationships connect the corresponding transaction entity nodes through edges; :A collection of information on policy and regulatory changes and market dynamics. Information on policy and regulatory changes and market dynamics related to carbon trading is also included as nodes in the relationship diagram, such as newly issued carbon emission standard policies, price fluctuation information in the carbon trading market, etc. These nodes can reflect the impact of external environmental factors on carbon trading relationships; : Node set, which is composed of the transaction subject node set and the policy and regulation change and market dynamic information node set, covering all node types in the carbon trading relationship diagram; : The edge set includes not only the transaction relationship edges between transaction entities, but also the association edges between transaction entities and policy and regulatory changes and market dynamic information. Through these edges, a complex relationship network between various elements in carbon trading is completely constructed.

[0071] Implementation steps: Collect information on trading entities in carbon trading and build a trading entity collection , determine the transaction relationship between transaction entities, and build a transaction relationship set , collect information on policy and regulatory changes and market dynamics related to carbon trading, and build a collection of information on policy and regulatory changes and market dynamics , build a node set and edge set , add the transaction relationship edges between transaction entities and the association edges between transaction entities and policy and regulatory changes and market dynamic information to the relationship graph to form a carbon trading relationship graph .

[0072] Operation logic: First, the trading subjects, trading relationships, and external policy and regulatory changes and market dynamics information in carbon trading are collected and sorted out respectively. Then, the trading subjects are regarded as one type of node, and policy and regulatory changes and market dynamics information are regarded as another type of node. Edges are constructed according to the trading relationships and the associations between them, thus forming a complete carbon trading relationship diagram. This relationship diagram can intuitively display the relationship between various elements in carbon trading, and provides a data structure foundation for the subsequent use of the GNN algorithm to mine abnormal patterns.

[0073] Beneficial effects: Adding basic trading entities, trading relationships, changes in carbon trading-related policies and regulations, and market dynamic information as virtual nodes and edges to the relationship graph can more comprehensively reflect the impact of the carbon trading environment on trading relationships, so that when mining abnormal patterns, not only the internal relationships between trading entities can be considered, but also external environmental factors can be comprehensively considered. This will help regulators grasp the overall health of the carbon trading market from a macro perspective, promptly discover systemic risks and illegal trading behaviors, and promote the standardization and sustainable development of the carbon trading market.

[0074] GNN Analysis and Verification Unit - GNN Algorithm Applied to Anomaly Pattern Mining Set up The node feature matrix of layer GNN is (Each row represents the feature vector of a node), the message passing function is , the aggregation function is , then the node feature update formula is:

[0075] in, Representation Node The set of neighbor nodes of is an activation function, such as ReLU function; In the process of information propagation, the message passing function can be expressed as:

[0076] in, For the The trainable weight matrix of the layer, , Node , In the The feature vector of the layer, To connect nodes and The eigenvector of the edge; Aggregation functions can be summed, averaged, and other methods. For example, when performing sum aggregation:

[0077]

[0078] in: : No. The node feature matrix of the layer GNN. In each layer of GNN, the features of the nodes are updated according to the information of their neighboring nodes. This matrix records the feature status of the nodes in each layer. : Message passing function, which generates messages to nodes based on the characteristics of the node itself, the characteristics of neighboring nodes, and the characteristics of the edges connecting them, through a trainable weight matrix These features are combined and transformed so that the node can obtain relevant information about neighboring nodes and update features based on this information; : Aggregation function is used to aggregate multiple messages received from neighboring nodes. Different aggregation methods (such as summation, average, etc.) will affect the node's comprehensive processing of neighbor information. Through the aggregation function, the node integrates the information of neighboring nodes into a single information representation for subsequent feature updates.

[0079] :node The neighbor node set of a node defines the node From which nodes information will be received, it reflects the local connection relationship between nodes in the graph structure; : Activation function, such as ReLU function , which performs nonlinear transformation on the aggregated and updated node features, increasing the expressive power of the model and enabling GNN to learn more complex relational patterns; : No. The trainable weight matrices of the layers are used to adjust the values ​​of these weight matrices during the training process of GNN, so that the message passing function can better capture the characteristic information of nodes and edges, thereby learning the abnormal patterns in the carbon trading relationship graph.

[0080] , :respectively nodes , In the The feature vectors of the layers are updated in each layer of the GNN according to the message passing and aggregation operations, reflecting the changes in the feature representation of nodes at different layers; : Connection point and The feature vector of the edge participates in the message passing process together with the node features, so that the information of the edge can also affect the update of the node features. In the carbon trading relationship graph, the edge may contain information such as the type and strength of the trading relationship, which may play an important role in mining abnormal patterns.

[0081] Implementation steps: Initialize the node feature matrix of each layer of GNN , can be set according to the initial attribute information of the nodes in the carbon trading relationship graph; For each layer , For the set number of GNN layers, for each node , according to its neighbor node set , using message passing function Count the messages received from neighbor nodes ( ; Using Aggregate Functions Aggregate the messages to get the aggregated information , update the formula according to the node characteristics Update node feature matrix ; go through After the information propagation of the layer and the node feature update, according to the final node feature matrix To judge abnormal patterns, for example, some abnormality detection rules can be set, such as when the node characteristic value exceeds a certain threshold range, when the relationship pattern between nodes is significantly different from the normal pattern, etc., and based on these rules, it can be judged whether there is an abnormal pattern in the carbon trading relationship diagram.

[0082] Operation logic: In the GNN analysis and verification unit, starting from the initial node feature matrix of the carbon trading relationship graph, in each layer of GNN, each node collects information from its neighbor nodes through the message passing function, and integrates this information using the aggregation function. Then, the node's own features are updated after the nonlinear transformation of the activation function. As the number of GNN layers increases, the node can gradually obtain the information of multi-hop neighbors, thereby learning a wider range of graph structure information and potential relationship patterns. Through the multi-layer GNN structure and different message passing functions and aggregation functions in different layers, and dynamically adjusting the scope and intensity of information propagation according to the node type and edge weight, the model can more accurately mine abnormal patterns in complex relationships. For example, for transaction subject nodes and policy and regulation change nodes, different message passing and aggregation strategies may be adopted to adapt to their different roles and importance in the carbon trading relationship graph. At the same time, the weight of the edge can reflect the closeness of the transaction relationship or the impact of policies and regulations on the transaction subject, thereby affecting the effect of information dissemination.

[0083] Beneficial effects: A multi-layer GNN structure is adopted, each layer of GNN uses different message passing functions and aggregation functions, and the scope and intensity of information dissemination are dynamically adjusted according to the node type and edge weight, which can more accurately mine abnormal patterns in complex relationships. It can not only mine the potential associations and abnormal patterns between trading entities, but also comprehensively consider the impact of external environmental factors (such as changes in policies and regulations, market dynamic information) on carbon trading relationships. This helps regulators to grasp the overall health of the carbon trading market from a macro level, promptly discover systemic risks and illegal trading behaviors, promote the standardization and sustainable development of the carbon trading market, and provide strong technical support for the supervision and risk prevention and control of carbon asset trading.

[0084] Interaction and security layer-blockchain and smart contract module Blockchain data storage and interaction unit-blockchain network data processing further includes the following: , the blocks in the blockchain are , each block contains a block header and block ; The transaction data is stored in the block body, that is, , indicating the The data in a block is a collection of transaction data A subset of The block header contains information such as the hash value of the previous block. , through this chain structure, the data cannot be tampered with and its integrity is guaranteed; In terms of data interaction, let the query request be , the query result is , then there exists a query function , indicating that according to the query request In blockchain Query and get the result ; in: :A collection of carbon asset trading data, which includes various carbon trading-related information, such as information about both parties to the transaction, transaction amount, carbon quota quantity, etc. These data are core data that need to be stored on the blockchain to ensure their security and traceability.

[0085] : A collection of blocks in a blockchain. A blockchain is a chain structure consisting of a series of blocks. Each block contains a certain amount of transaction data and related block header information.

[0086] : A collection of block headers, which store some key information, such as the hash value of the previous block. The hash function is one-way and collision-resistant. By performing a hash operation on the information of the previous block, the hash value is stored in the current block header, so that the blocks in the blockchain form a chain structure. Once the data in a block is tampered with, the hash value of the subsequent blocks will not match, thus ensuring the data cannot be tampered with and its integrity.

[0087] : A block collection is used to store transaction data. The data in each block is part of the transaction data collection. Distributing transaction data in different blocks facilitates data management and query, while also ensuring data security.

[0088] : Hash function, which converts input data (such as block data) into a hash value of fixed length. In blockchain, hash functions play a key role, such as generating the hash value of the previous block in the block header, hashing transaction data to verify data integrity, etc.

[0089] and : Represents query request and query result respectively. In the blockchain data storage and interaction unit, a query interface needs to be provided so that users or other systems can query the carbon trading data stored on the blockchain. The query function , indicating that according to the given query request , in blockchain Data is retrieved and processed in the query, and finally the query results are obtained .

[0090] Implementation steps: When new carbon asset transaction data is generated, it is organized into a transaction data set Transaction data elements in , package transaction data into blocks according to certain rules , while calculating the block header The generated block is added to the blockchain using information from the previous block, such as the hash value of the previous block. In the process, a new blockchain state is formed. When a query request is received When, according to the query function , retrieve and process data in the blockchain, find transaction data related to the query request, and organize it into query results Return to the querying party.

[0091] Operation logic: New carbon trading data is first organized into a set of transaction data, then packaged into blocks together with block header information and added to the blockchain. During the data storage process, hash functions are used to ensure the chain structure between blocks and the immutability of data. During data interaction, each block in the blockchain is traversed according to the query request, and the transaction data that meets the requirements is found and the results are returned. During the whole process, the blockchain network, as a distributed ledger, ensures the security, traceability and immutability of carbon trading data, and provides a reliable data storage and interaction basis for carbon asset trading.

[0092] Beneficial effects: The establishment of a blockchain network to store key data and provide a query interface ensures the security and traceability of carbon asset trading data. The immutable nature of the data effectively prevents the transaction data from being maliciously tampered with or forged, providing a guarantee for the fairness and justice of carbon trading. At the same time, traceability enables the transaction history to be clearly queried and audited, which helps regulators to supervise and conduct compliance checks on the carbon trading market, and promotes the healthy and orderly development of the carbon trading market.

[0093] Smart contract execution unit-smart contract function implementation Assume that the carbon data submission function is , for the carbon data submitted; The verification process function is ; The transaction settlement function is , For transaction information; The violation penalty function is , For illegal information; A smart contract can be represented as: is established, the corresponding function is executed, for example, if Representing Carbon Data Verification passed and transaction information If legal, execute the transaction settlement function .

[0094] in: : Carbon data submission function, used to receive carbon data submitted externally ,These data may come from the data collection layer or other related systems, and are the basic data source for subsequent processing of smart contracts; :Verification process function, submit carbon data Verification operations may involve the invocation of various verification algorithms and models in the verification model layer mentioned above to determine whether the carbon data meets the requirements, such as the authenticity, accuracy, and completeness of the data.

[0095] :Transaction settlement function, when carbon data verification is passed and transaction information When legal, this function is executed to process the settlement operations of carbon asset transactions, such as the transfer of funds and carbon quotas, to ensure the smooth completion of transactions; : Violation penalty function, when violation information is detected , such as carbon data fraud, illegal trading, etc., this function is executed, and the violators are punished accordingly according to the preset violation penalty rules, such as fines, trading restrictions, etc., to maintain the order of the carbon trading market; :Smart contract, which is a collection of the above-mentioned core functions, defines the main operating logic in the carbon asset trading and verification process.

[0096] :Conditional function, used to determine whether the conditions for executing a smart contract function are met. These conditions are usually related to carbon data, transaction information and violation information. The execution of the corresponding smart contract function is triggered by judging these information, ensuring that the execution of the smart contract complies with the preset rules and logic.

[0097] Implementation steps: External system calls carbon data submission , the carbon data Submit to the smart contract, and the smart contract calls the verification process function internally Verify the submitted carbon data and and violation information , judgment condition function Is it true? If the conditions are met, such as for the transaction settlement function, when the carbon data verification is passed and the transaction information is legal, the transaction settlement function is executed For the violation penalty function, when violation information is detected, the violation penalty function is executed .

[0098] Operation logic: Smart contracts operate according to pre-set rules and logic. They first receive carbon data submitted externally and then conduct verification. Based on the verification results, transaction information, and violation information, they determine whether the conditions for executing a specific function are met. If the conditions are met, the corresponding function is executed to complete operations such as carbon data processing, transaction settlement, or violation penalties. The entire process is automatically executed without human intervention, ensuring the efficiency and accuracy of the carbon asset trading and verification process.

[0099] Beneficial effects: Write smart contract code and automatically execute carbon asset trading and verification-related operations based on conditions. Use modular design to encapsulate carbon data submission, verification process, transaction settlement, violation penalties and other functions into independent modules, which is convenient for contract maintenance and upgrades. When carbon trading rules or verification processes change, only the corresponding modules need to be modified or upgraded in a targeted manner, without the need to make large-scale changes to the entire smart contract, which greatly reduces the complexity and cost of contract maintenance. At the same time, setting up an audit mechanism for smart contracts allows regulators and third-party auditing agencies to supervise and audit the contract execution process, further enhancing the transparency and credibility of smart contracts, ensuring that carbon asset trading and verification processes strictly follow preset rules, effectively preventing loopholes in the contract execution process from being exploited, improving the security and credibility of the entire carbon trading system, and promoting the healthy and orderly operation of the carbon asset trading market. Specific embodiment 4

[0101] like Figure 1-3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed: The operating steps of the entire system are as follows: The steps of data transmission in each layer of the whole system are as follows: Sp1: Data collection: In the data collection layer, the federated learning subunit prompts each data provider to train the local model and encrypt and upload the parameters to the coordination server, which aggregates and updates to train the global model; the edge computing subunit performs adaptive compression and other pre-processing on the raw data at the data source end, and transmits suspicious abnormal data to the data integration layer; Sp2: Data integration: After the data integration layer receives the data, the ontology construction unit constructs the carbon trading ontology model; the semantic mapping and association unit integrates the data according to rules and machine learning methods; the data storage and management unit stores and manages data permissions; Sp3: Data verification: Verify the model layer work, integrate the LSTM-SVM verification module, the LSTM unit analyzes the historical data to obtain the feature vector, and the SVM unit judges whether the data is normal or not based on it; in the GNN relationship verification module, first build the carbon trading relationship diagram, and then the GNN unit mines the abnormal pattern; Sp4: Interaction and Execution: Interaction and security layer actions, the blockchain data storage and interaction unit stores and verifies key data and provides queries; the smart contract execution unit writes code based on the data and automatically executes transactions and verification operations, and outputs the results.

[0102] Through these four steps, the system completes the entire process of carbon asset trading data processing, ensures the accuracy and security of transaction data verification, and promotes orderly transactions.

[0103] Sp1: Data collection layer: multi-source data acquisition and preliminary processing: The data collection layer of the system first obtains carbon asset trading related data from multiple sources. These data sources are diverse and cover all aspects related to carbon asset trading; The federated learning and edge computing modules in this layer play an important role. The federated learning subunit establishes a federated learning coordination server to communicate and coordinate with the local nodes of each data provider. Each data provider trains a local model based on local carbon data, encrypts the local model parameters using homomorphic encryption technology, and uploads the encrypted model parameters to the coordination server. The coordination server then uses a multi-party computing protocol to aggregate the encrypted parameters and update the local model to iteratively train the global model.

[0104] At the same time, the edge computing sub-unit deploys edge computing devices near the internal data source of the enterprise to perform real-time preprocessing of the original data. During the preprocessing process, an adaptive compression algorithm is used according to the time series characteristics and data type of the data. For example, a higher compression ratio is used for basic data that changes slowly, and a lower compression ratio is used for real-time monitoring data with large fluctuations. After preprocessing, the data is compressed, format converted and preliminarily screened, and the screened suspicious abnormal data is transmitted to the data integration layer.

[0105] Sp2: Data integration layer: integration association and storage management based on semantic web technology: After the data integration layer receives the suspicious abnormal data filtered by the data collection layer, it processes it through the semantic web data integration module; First, the ontology construction unit builds the ontology model of the carbon trading field based on the internationally accepted carbon trading standards and industry specifications, using a combination of top-down and bottom-up construction methods. It first determines the core concept framework, then gradually refines the sub-concepts and attributes, and regularly updates and maintains the ontology model according to changes in policies and regulations and industry development trends. Next, the semantic mapping and association unit maps and matches the collected data with the ontology model and associates and integrates them. When performing semantic mapping, a mapping method combining rule-based and machine learning-based methods is adopted. Predefined rules are used for rapid mapping of common data formats and standard terms, and machine learning models are used for automatic learning and mapping of fuzzy data expressions. Finally, the data storage and management unit classifies and stores the integrated data and manages access rights, providing an orderly and manageable data foundation for subsequent verification processes.

[0106] Sp3: Verification model layer: Verification using multiple algorithm models: The verification model layer conducts verification work on the integrated data, mainly by integrating the LSTM-SVM verification module and the GNN relationship verification module; In the LSTM-SVM verification module, the LSTM time series analysis unit first obtains historical carbon data and related time series data, and uses the sliding window technology to segment the historical carbon data to increase the number of training samples. The early stopping method is combined to prevent overfitting, and the attention mechanism is introduced to enable the model to focus on the time series feature parts that have an important impact on carbon verification. Then, the LSTM network model is trained to output the time series feature vector and the predicted trend value. Then, the SVM classification verification unit receives the feature vector, uses the kernel function technique, selects the radial basis function as the kernel function, maps the nonlinearly separable data, and optimizes the kernel function parameters and penalty factors through the cross-validation method. The trained SVM model is used to determine whether the data is normal. In terms of the GNN relationship verification module, the carbon trading relationship graph construction unit adds basic trading entities, trading relationships, carbon trading-related policy and regulatory changes, and market dynamic information as virtual nodes and edges into the relationship graph to construct a carbon trading relationship graph. Afterwards, the GNN analysis and verification unit uses a multi-layer GNN structure based on this relationship graph. Each layer of GNN uses different message passing functions and aggregation functions, and dynamically adjusts the scope and intensity of information dissemination according to the node type and edge weight, so as to more accurately mine abnormal patterns in complex relationships. Sp4: Interaction and security layer: Ensure data interaction is secure and reliable and execute transaction and verification processes: The blockchain and smart contract modules in the interaction and security layer are responsible for ensuring the data interaction security of the entire system and automating the transaction and verification process; The blockchain data storage and interaction unit establishes a blockchain network to store key data and provides a query interface to ensure data storage security and queryability; The smart contract execution unit adopts a modular design, encapsulating functions such as carbon data submission, verification process, transaction settlement, and violation penalties into independent modules. When the carbon trading rules or verification process change, only the corresponding modules need to be modified or upgraded in a targeted manner. At the same time, an audit mechanism for smart contracts is set up to allow regulators and third-party auditing agencies to supervise and audit the contract execution process, ensure that the carbon asset trading and verification process strictly follows the preset rules, ensure the security and credibility of the entire carbon asset trading system, and promote the healthy and orderly operation of the carbon asset trading market.

[0107] Through the collaborative operation of the above layers, the system realizes a complete data processing route from data collection, integration, verification to ensuring the security of data interaction and executing transaction verification processes, effectively improving the accuracy and reliability of data verification in carbon asset transactions and the security and credibility of the entire trading system. Specific embodiment five

[0109] like Figure 1-3As shown, according to the contents in the above specific embodiments, the following contents are further disclosed: When the entire system is actually used, the hardware architecture corresponding to the modules and units at each level is as follows: Hardware description of the federated learning and edge computing modules of the data collection layer: Federated learning subunit hardware requirements: Server equipment: A high-performance server needs to be configured to act as a federated learning coordination server to communicate and coordinate with the local nodes of each data provider and process operations such as aggregation parameters. The server should have high processing capabilities, such as multi-core processors, such as Intel Xeon series, to cope with the operation of large amounts of data and communication coordination between multiple nodes. At the same time, it should be equipped with large-capacity memory, such as 32GB and above, to ensure the smoothness of the data processing process, and high-speed network interfaces, such as 10 Gigabit Ethernet interfaces, to ensure fast and stable data transmission with the nodes of each data provider; Data provider local node equipment: The local nodes of each data provider may be various types of computer equipment, such as desktop computers or servers. For desktop computers, they should generally have medium-performance processors, such as Intel Core i5 and above, an appropriate amount of memory, such as 8GB and above, and a stable network connection, such as a Gigabit Ethernet interface, to ensure that the model can be trained based on local carbon data, and to achieve communication with the coordination server, upload encrypted model parameters, and other operations. If a server is used as a local node, its hardware configuration can refer to the low-profile version of the federated learning coordination server, such as a multi-core processor, a lower-configuration model of the Intel Xeon series, 16GB and above memory, etc. It also needs to have a high-speed network interface to ensure data transmission; Edge computing subunit hardware requirements: Edge computing devices: Edge computing devices deployed near the internal data source of an enterprise are key hardware for real-time preprocessing of raw data. Such devices can be dedicated edge computing servers or industrial gateways with edge computing functions. Taking edge computing servers as an example, they need to have certain processing capabilities. They can usually use low-power but moderate-performance processors, such as Intel Atom series or ARM architecture processors, and be equipped with an appropriate amount of memory, such as 4GB and above, to meet the operating requirements of operations such as data compression, format conversion and preliminary screening. At the same time, it should have multiple data interfaces, such as common Ethernet interfaces for connecting to the internal network of the enterprise, and possible serial ports and other interfaces for connecting different types of data source devices, such as sensors, so as to obtain raw data for preprocessing. In addition, considering that it may be deployed in some industrial environments, the equipment should also have certain dust, moisture, and earthquake protection performance.

[0110] Hardware related description of the semantic web data integration module of the data integration layer: This module mainly focuses on the software-level processing of data, such as building ontology models, semantic mapping and association, and data storage and management. Relatively speaking, the dependence on specific hardware devices is not as direct as that of the data acquisition layer, but in actual operation, it still needs to rely on a certain hardware foundation; Server equipment: Generally, one or more servers are required to run the software programs related to the semantic web data integration module. The server can be a general-purpose server with moderate performance, such as an Intel Core i7 series processor, 16GB or more memory, and a large-capacity hard disk, such as 1TB or more, to store the integrated data and run related software programs. The server should have a high-speed network interface, such as a Gigabit Ethernet interface, to ensure the data transmission speed between other layers such as the data collection layer and the verification model layer.

[0111] Hardware related instructions for the fused LSTM-SVM verification module and GNN relationship verification module of the verification model layer: High-performance computing equipment: Since these two verification modules involve relatively complex algorithm operations, such as LSTM network model training, SVM model processing, and GNN algorithm mining, high-performance computing equipment is required to ensure the speed and accuracy of the operation. Usually, a dedicated high-performance server or workstation can be used, for example, equipped with a multi-core high-performance processor, such as Intel Xeon series high-performance models, large-capacity memory, such as 64GB and above, and high-speed storage devices, such as solid-state drives. The capacity can be determined according to actual needs, and 500GB and above is generally recommended for storing training data, model files, etc. At the same time, these devices should have high-speed network interfaces, such as 10 Gigabit Ethernet interfaces, to ensure the rapid transmission of data between different modules and ensure the efficient implementation of the verification process.

[0112] Blockchain and smart contract module hardware description for the interaction and security layer: Blockchain node devices: For blockchain data storage and interaction units, if you want to establish a blockchain network to store key data and provide a query interface, you need to deploy blockchain node devices. These devices can be dedicated blockchain servers or general servers with blockchain functions. Taking blockchain servers as an example, their hardware configuration can be similar to high-performance computing devices, that is, equipped with multi-core high-performance processors, such as Intel Xeon series high-performance models, large-capacity memory, such as 64GB and above, and high-speed storage devices, such as solid-state drives. The capacity can be determined according to actual needs, and 500GB and above is generally recommended. At the same time, in order to ensure the stable operation of the blockchain network and the security of data transmission, it is necessary to have a high-speed and secure network interface, such as a 10 Gigabit Ethernet interface and cooperate with encrypted transmission protocols.

[0113] Smart contract execution equipment: When the smart contract execution unit performs related operations, it also relies on certain hardware equipment. Generally, the software programs related to smart contracts can be run on the above-mentioned servers or workstations. Its hardware configuration requirements are similar to those of the equipment running the verification model layer related software programs, that is, equipped with multi-core high-performance processors, such as Intel Xeon series high-performance models, large-capacity memory, such as 64GB and above, and high-speed storage devices, such as solid-state drives. The capacity can be determined according to actual needs, and 500GB and above is generally recommended. This can ensure the efficiency and accuracy of smart contracts in executing carbon data submission, verification process, transaction settlement, violation penalties and other functions.

[0114] The above hardware is only a rough reference based on the system function requirements. In actual application scenarios, the hardware equipment can be further optimized and adjusted according to specific business scale, data volume, performance requirements and other factors. Specific embodiment six

[0116] like Figure 1-3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed: When the entire system is actually used, when facing massive carbon asset trading data, it ensures the real-time and completeness of data collection, and further includes the following contents: Data Collection Layer Federated learning and edge computing modules: Federated learning subunit: Establish a federated learning coordination server to communicate and coordinate with the local nodes of each data provider. Each data provider trains a local model based on local carbon data and encrypts and uploads the model parameters to the coordination server. The server aggregates the parameters and sends the updated local model to iteratively train the global model. This distributed learning method can make full use of the data resources of all parties. Without directly sharing the original data, it can realize the collaborative training of the model, ensure the integrity of data collection, and ensure the timeliness of data updates through an efficient communication coordination mechanism; Edge computing subunit: Edge computing devices are deployed near the enterprise's internal data sources to perform real-time preprocessing of raw data, including data compression, format conversion, and preliminary screening. During the data compression process, an adaptive compression algorithm is used based on the time series characteristics and data type of the data. A higher compression ratio is used for basic data that changes slowly, and a lower compression ratio is used for real-time monitoring data with large fluctuations. While ensuring data validity, the amount of data transmission is minimized to the greatest extent, thereby improving the real-time nature of data transmission. The screened suspicious abnormal data is transmitted to the data integration layer to ensure data integrity. Adopting RPA robots and other technical means: For example, the carbon asset management RPA robot independently developed by Jiangsu Petroleum can read the basic information and related data of gas stations, match the basic data with the reporting items one by one, automatically calculate and fill the results into the carbon asset system form template, and realize the automation of operations such as data reading, reporting item matching, accounting and deduction. It can complete large amounts of data processing in a short time, effectively avoid problems such as incorrect filling, missing filling, and calculation errors, provide protection for the integrity of data collection, and its high efficiency helps to improve the real-time performance of data collection.

[0117] Data Integration Layer Semantic Web Data Integration Module: Ontology construction unit: Construct an ontology model in the field of carbon trading, define core concepts and their interrelationships, adopt a top-down and bottom-up construction method based on internationally accepted carbon trading standards and industry specifications, and regularly update and maintain the ontology model according to changes in policies and regulations and industry development trends. This enables the collected data to be integrated according to unified standards and specifications, ensuring the consistency and integrity of the data at the semantic level, laying the foundation for subsequent accurate verification and analysis; Semantic mapping and association unit: A mapping method combining rule-based and machine learning is used to map and match the collected data with the ontology model and associate and integrate them. Predefined rules are used to quickly map common data formats and standard terms. Ambiguous data expressions are automatically learned and mapped using machine learning models, so as to comprehensively and accurately integrate various types of data and ensure the integrity and accuracy of data integration. Data storage and management unit: Classify and store integrated data and manage access rights. Through reasonable data classification and strict access control, ensure orderly storage and secure access to data, prevent data loss or tampering, and further ensure data integrity.

[0118] Interaction and security layer Blockchain and smart contract module: Blockchain data storage and interaction unit: Establish a blockchain network to store key data and provide a query interface. The multi-party participation, non-tamperability, and traceability of blockchain technology ensure the authenticity and integrity of the data. Once the data is on the chain, it can be shared and queried in real time, ensuring the real-time and transparency of the data, and providing a reliable foundation for the collection and management of carbon asset trading data; Smart contract execution unit: Write smart contract code and automatically execute carbon asset trading and verification-related operations according to conditions. Adopt modular design to encapsulate carbon data submission, verification process, transaction settlement, violation penalties and other functions into independent modules, which is convenient for contract maintenance and upgrade. At the same time, set up an audit mechanism for smart contracts, allowing regulatory agencies and third-party audit institutions to supervise and audit the contract execution process, ensure that data collection and transaction processes are strictly carried out in accordance with preset rules, and ensure data integrity and compliance.

[0119] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprising a reference structure" do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0120] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An automated data verification system for carbon asset trading, characterized in that: The system includes a data collection layer, a data integration layer, a verification model layer, and an interaction and security layer, wherein: The data acquisition layer is used to obtain carbon asset trading related data from multiple sources and perform preliminary processing, including a federated learning and edge computing module, which is used to realize the collaborative training and verification model of multiple data holders through a federated learning algorithm and pre-process data at the data source end using edge computing. The federated learning and edge computing module includes a federated learning subunit and an edge computing subunit. The federated learning subunit is used to establish a federated learning coordination server to communicate and coordinate with the local nodes of each data provider. Each data provider trains a local model based on local carbon data and encrypts and uploads the model parameters to the coordination server. After the server aggregates the parameters, it sends down an updated local model to iteratively train the global model. The edge computing subunit is used to deploy edge computing equipment near the internal data source of the enterprise to perform real-time pre-processing on the raw data, including data compression, format conversion and preliminary screening. The screened suspicious abnormal data is transmitted to the data integration layer; The data integration layer is used to integrate, associate, store and manage the collected data based on semantic web technology, including a semantic web data integration module, which includes an ontology construction unit, a semantic mapping and association unit, and a data storage and management unit. The ontology construction unit is used to construct an ontology model in the field of carbon trading to define core concepts and their interrelationships. The semantic mapping and association unit is used to map and match the collected data with the ontology model and associate and integrate them. The data storage and management unit is used to classify and store the integrated data and manage access rights. The verification model layer is used to verify the integrated data using multiple algorithm models, including a fused LSTM-SVM verification module and a GNN relationship verification module. The fused LSTM-SVM verification module includes an LSTM time series analysis unit and an SVM classification verification unit. The LSTM time series analysis unit is used to obtain historical carbon data and related time series data to train the LSTM network model to output a time series feature vector and a predicted trend value. The SVM classification verification unit is used to receive the feature vector and determine whether the data is normal based on the trained SVM model. The GNN relationship verification module includes a carbon trading relationship graph construction unit and a GNN analysis and verification unit. The carbon trading relationship graph construction unit is used to construct a carbon trading relationship graph. The GNN analysis and verification unit is used to mine abnormal patterns based on the relationship graph using the GNN algorithm; The interaction and security layer is used to ensure the security and reliability of data interaction and automatically execute transactions and verification processes, including a blockchain and smart contract module. The blockchain and smart contract module includes a blockchain data storage and interaction unit and a smart contract execution unit. The blockchain data storage and interaction unit is used to establish a blockchain network to store key data and provide a query interface. The smart contract execution unit is used to write smart contract code and automatically execute carbon asset trading and verification related operations based on conditions.

2. The automated data verification system in carbon asset trading according to claim 1, characterized in that: In the federated learning subunit, each data provider's local node uses homomorphic encryption technology to encrypt local model parameters to ensure that data privacy is not leaked during the parameter upload process, and the coordination server uses a multi-party computing protocol to aggregate the encrypted parameters to achieve secure global model parameter updates.

3. The automated data verification system in carbon asset trading according to claim 1, characterized in that: During the data compression process, the edge computing subunit adopts an adaptive compression algorithm according to the time series characteristics and data type of the data, adopts a higher compression ratio for basic data that changes slowly, and adopts a lower compression ratio for real-time monitoring data that fluctuates greatly, thereby minimizing the amount of data transmission while ensuring data validity.

4. The automated data verification system in carbon asset trading according to claim 1, characterized in that: When constructing the ontology model in the field of carbon trading, the ontology construction unit adopts a top-down and bottom-up construction method based on internationally accepted carbon trading standards and industry specifications, first determining the core concept framework, then gradually refining the sub-concepts and attributes, and regularly updating and maintaining the ontology model according to changes in policies and regulations and industry development trends.

5. The automated data verification system in carbon asset trading according to claim 1, characterized in that: The semantic mapping and association unit adopts a mapping method that combines rule-based and machine learning-based mapping when performing semantic mapping. Predefined rules are used to quickly map common data formats and standard terms, and machine learning models are used to automatically learn and map fuzzy data expressions.

6. The automated data verification system in carbon asset trading according to claim 1, characterized in that: When training the LSTM network model, the LSTM time series analysis unit uses sliding window technology to segment the historical carbon data to increase the number of training samples, and combines the early stopping method to prevent overfitting. At the same time, an attention mechanism is introduced to enable the model to focus on the time series feature parts that have an important impact on carbon verification.

7. The automated data verification system in carbon asset trading according to claim 1, characterized in that: When constructing the classification hyperplane, the SVM classification verification unit uses the kernel function technique to map the nonlinearly separable data, selects the radial basis function as the kernel function, and optimizes the kernel function parameters and penalty factors through the cross-validation method to improve the accuracy of the SVM model in classifying complex carbon data.

8. The automated data verification system in carbon asset trading according to claim 1, characterized in that: When constructing the carbon trading relationship graph, the carbon trading relationship graph construction unit adds basic trading entities, trading relationships, changes in carbon trading-related policies and regulations, and market dynamic information into the relationship graph as virtual nodes and edges to more comprehensively reflect the impact of the carbon trading environment on trading relationships.

9. The automated data verification system in carbon asset trading according to claim 1, characterized in that: The GNN analysis and verification unit adopts a multi-layer GNN structure when performing information propagation and node feature update. Each layer of GNN adopts different message passing functions and aggregation functions, and dynamically adjusts the scope and intensity of information propagation according to the node type and edge weight, so as to more accurately mine abnormal patterns in complex relationships.

10. The automated data verification system in carbon asset trading according to claim 1, characterized in that: When writing the smart contract code, the smart contract execution unit adopts a modular design, encapsulating the functions of carbon data submission, verification process, transaction settlement, and violation penalty into independent modules to facilitate contract maintenance and upgrades. At the same time, an audit mechanism for smart contracts is set up to allow regulators and third-party audit agencies to supervise and audit the contract execution process.

Images