Blockchain-based Trusted Management Method for Supply Chain Carbon Data

Abstract

The present invention relates to the technical field of data processing, and relates to a method for trusted management of supply chain carbon data based on blockchain. By deploying an edge computing gateway, the energy consumption of suppliers' production equipment is collected in real time. Among them, the suppliers' production data collects the energy consumption in real time through the edge computing gateway deployed on the suppliers' production equipment, combines the on-vehicle terminal to obtain the real-time operating parameters of the logistics transportation vehicle, integrates the material disassembly data in the recycling and processing link, and constructs a multi-source data collection system for the entire supply chain link. The blockchain network performs trusted processing on the data, calls the regional carbon emission factors in the industry benchmark database for deviation verification, and the intelligent contract engine executes audit decisions based on preset compliance rules. When it is detected that the energy consumption deviates from the benchmark threshold, a data correction process confirmed by multi-party signatures is triggered. The present invention realizes the trusted storage and dynamic optimization of carbon data throughout the entire link, and improves the accuracy of supply chain carbon emission accounting and the efficiency of cross-border audits.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a method for trustworthy management of supply chain carbon data based on blockchain. Background Art

[0002] There are technical bottlenecks in the existing supply chain carbon data management, such as data silos in multiple links and the lack of a verification mechanism, resulting in insufficient accuracy of carbon emission accounting throughout the life cycle. The existing methods rely on manual collection of scattered supplier energy consumption data and static transportation parameters, making it difficult to integrate heterogeneous data sources such as raw material procurement traceability information, energy efficiency parameters of production equipment, and dynamic working conditions of logistics carriers in real time, resulting in calculation deviations of carbon emission factors. The data storage formats of each node in the supply chain are not unified, and there is a lack of a cross-link traceability and verification mechanism, making it impossible to effectively verify the correlation between supplier life cycle assessment data and actual production energy consumption. In the carbon emission accounting of logistics transportation, fixed emission factors are mostly used, and no dynamic correction is made in combination with real-time route optimization and load changes. The audit node algorithm lacks a collaborative verification mechanism for abnormal detection of transportation trajectories and analysis of energy consumption fluctuations in warehouses. The application of blockchain technology is mostly limited to data storage and certification, and has not solved technical problems such as standard integration of multi-source data, dynamic monitoring and early warning, and adaptation of cross-border audit rules, restricting the full-link trustworthy control of the supply chain carbon footprint. Summary of the Invention

[0003] Aiming at the deficiencies of the existing technology, the present invention provides a method for trustworthy management of supply chain carbon data based on blockchain, and the technical problems to be solved are the problems of low carbon emission accounting accuracy, insufficient audit efficiency, and lack of a trustworthy traceability system caused by difficulties in integrating multi-source heterogeneous data, lack of cross-link traceability and verification, and imperfect dynamic monitoring mechanism in the existing supply chain carbon data management.

[0004] To solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0005] The method for trustworthy management of supply chain carbon data based on blockchain provided by the present invention includes:

[0006] Step S101, obtaining data of all links in the supply chain. The data of all links in the supply chain includes supplier production data, manufacturing process data, logistics transportation data, and recycling and processing data. Among them, the supplier production data is used to collect the energy consumption in real time through an edge computing gateway deployed on the supplier production equipment. The manufacturing process data includes energy equipment sensor data and process improvement records in the factory. The logistics transportation data is used to obtain the real-time working condition parameters of the transportation vehicle through an on-vehicle terminal. The recycling and processing data includes the recovery rate of disassembled materials of the product.

[0007] Step S102: Input the data of all links in the supply chain into the blockchain network for trust processing. Generate a data digest including spatio-temporal characteristics through the data standardization module, and write the data digest into the blockchain distributed ledger. Among them, the following processing is performed on the supplier production data: Call the regional carbon emission factors in the industry benchmark database to perform data deviation verification, and generate an on-chain record including a verification identifier; Combine the manufacturing process data with the equipment operation parameters to construct a dynamic emission factor model, and verify the logical relevance between the process improvement record and the energy consumption through the dynamic emission factor model.

[0008] Step S103: Input the on-chain record into the intelligent contract execution engine to make an audit decision based on the preset carbon emission compliance rules. When it is detected that the verification identifier corresponding to the energy consumption deviates from the industry benchmark threshold, trigger the data correction instruction generation process, and perform multi-party signature confirmation through the blockchain consensus node.

[0009] Step S104: Generate a supply chain optimization instruction according to the result of the audit decision, and feedback the supply chain optimization instruction to the management-side execution device. The supply chain optimization instruction includes supplier evaluation parameters and transportation path planning strategies dynamically adjusted based on the anomaly detection result, forming a low-carbon supply chain management closed-loop system that runs through data collection, verification, and decision-making.

[0010] Collect power consumption data at a set sampling frequency through the edge computing gateway deployed on the supplier production equipment, and realize real-time acquisition of multi-source data in combination with the industrial Internet of Things protocol. The data standardization module calls the carbon emission factors in the industry benchmark database to convert the original energy consumption into the standard carbon emission equivalent value. The dynamic emission factor model dynamically adjusts the calculation logic according to the equipment operation state parameters. When it is detected that there is a significant deviation between the process improvement record and the energy consumption change, generate an on-chain record with a verification identifier. The intelligent contract engine triggers the data correction process based on the preset rules, and the audit node uses an encryption algorithm to perform multi-party verification on the correction instruction, and finally generates a supplier evaluation parameter adjustment instruction and feedbacks it to the management side.

[0011] Further, in the method for trust management of supply chain carbon data based on blockchain of the present invention, step S102 further includes:

[0012] Call the raw material procurement database of the third-party certification agency through the blockchain oracle interface, extract the hash digest of the original procurement voucher, and generate a hash feature value including a timestamp;

[0013] Input the hash feature value and the bill of materials data in the life cycle assessment report into the zero-knowledge proof verification module to verify the relevance and consistency between the procurement voucher and the carbon emission inventory;

[0014] Generate a blockchain data block including a trusted identifier based on the verification result, where the trusted identifier is used to update the evaluation status of the integrity of supplier production data.

[0015] During the processing of supplier production data, the blockchain oracle interface accesses the procurement database of the polysilicon supplier in Region C, performs the SHA-256 hash operation on the procurement order to generate a 128-bit eigenvalue. The zero-knowledge proof module verifies the correlation between the eigenvalue and the LCA report provided by the certification agency in Region D. When the energy consumption deviation exceeds the threshold, a data block with a trusted identifier is generated. This identifier triggers the supplier status update process in the supply chain management system, realizing the dynamic matching of procurement traceability data and actual production energy consumption.

[0016] Furthermore, for the blockchain-based trusted management method of supply chain carbon data according to the present invention, the step S102 includes

[0017] Access the waste disposal blockchain of the regulatory platform through a cross-chain relay node to obtain digitally signed electronic consignment note data;

[0018] Extract the time-series consumption data of the chemical inventory management system at the production end, and calculate the dynamic deviation rate between the time-series consumption data and the disposal volume of the electronic consignment note;

[0019] When the dynamic deviation rate exceeds the preset threshold, trigger a smart contract to lock the on-chain status of the carbon emission calculation result for the corresponding production batch.

[0020] The verification of manufacturing process compliance is implemented in the chemical plant in Region E. The cross-chain relay node accesses the hazardous waste disposal blockchain of the Department of Ecology and Environment in Region F to obtain electronically signed consignment note data using the national cryptographic algorithm. The chemical inventory management system synchronizes the consumption data every 15 minutes. When it is detected that the monthly consumption deviation of dichloromethane reaches 4.3% of the disposal volume, the smart contract automatically locks the carbon emission calculation result for this batch. The dynamic emission factor model generates a correction coefficient by combining the boiler thermal efficiency parameter of 78%. The boiler thermal efficiency of 78% represents the proportion of effective thermal energy in the total fuel energy during the fuel combustion process. The higher the thermal efficiency, the higher the actual energy utilization rate per unit of fuel, the lower the fuel consumption under the same output, and the corresponding carbon emissions should be less. The dynamic emission factor model dynamically corrects the industry benchmark emission factor (based on the ideal thermal efficiency or industry average) by introducing the real-time boiler thermal efficiency parameter to reflect the impact of the actual operating state of the equipment on carbon emissions, writes it into the Hyperledger Fabric chain code and triggers an audit alarm.

[0021] Furthermore, for the blockchain-based trusted management method of supply chain carbon data according to the present invention, the step S102 includes:

[0022] Based on the starting or ending coordinates of the transportation task and the vehicle load parameters, call the path planning algorithm to generate the benchmark carbon emission value of the theoretically optimal transportation path;

[0023] Obtain the real-time working condition data of the actual transportation trajectory through on-vehicle sensors, and calculate the incremental difference between the actual carbon emissions and the benchmark carbon emission value;

[0024] When the incremental difference exceeds the preset path deviation threshold in the smart contract, create an abnormal transportation event record in the blockchain ledger and trigger a suspension execution instruction for the on-chain freight settlement contract.

[0025] The logistics anomaly detection is applied to the transportation task of auto parts from Region G to Region H. The path planning engine generates a benchmark carbon emission value of 2.1 tons of CO2 based on parameters of a load of 28 tons and a distance of 850 kilometers. The actual path deviation of 37 kilometers collected in real time by the on-vehicle OBD terminal triggers a smart contract warning. The blockchain ledger creates an abnormal event record including the transportation number GL2024-07, and synchronously freezes the on-chain freight settlement contract. The edge computing node calls the historical load data to verify the transportation rationality, and the corrected data is synchronized to the regulatory platform in Region I through the cross-chain protocol.

[0026] Further, the method for trusted management of supply chain carbon data based on blockchain according to the present invention further includes: parsing the energy consumption log data of port handling equipment through an edge computing node, and extracting the speed and cargo capacity parameters in the ship AIS trajectory data;

[0027] Synchronously collect the GPS trajectory data and load change parameters of road transport vehicles, the freight formation information and energy consumption monitoring data of railway transport;

[0028] Build a carbon emission calculation model for the sea transportation segment based on the speed and cargo capacity parameters, and integrate the real-time load, path optimization parameters of road transport and the energy consumption coefficient of railway transport to generate a carbon emission integration model covering multimodal transport;

[0029] Extract the carbon emission intensity characteristics of sea, road and railway transport, and map them to the transportation identifier through a preset coding rule;

[0030] Generate a carbon emission feature vector including ship identifiers, road transport batch codes and railway freight waybills;

[0031] Cross-chain hash bind the carbon emission feature vector with the blockchain data of the road, railway and sea transportation segments, and generate a carbon footprint tracking map covering multiple transportation modes of road, railway and sea through a smart contract.

[0032] Further, in the method for trusted management of supply chain carbon data based on blockchain according to the present invention, the step S103 includes:

[0033] Based on the metal element content parameters marked in the product design bill of materials, calculate the theoretical recovery threshold according to the metallurgical recovery rate formula;

[0034] Input the actual extraction quantity reported by the recycler and the theoretical recovery threshold into the blockchain smart contract for deviation comparison;

[0035] When it is detected that the actual extraction quantity exceeds the theoretical threshold and the continuous duration exceeds the preset period, activate the supplier performance deduction rule module in the smart contract, and send an alarm instruction including the abnormal batch identifier to the blockchain node with supervision authority.

[0036] The recycling process audit is carried out in the battery recycling plant in region M. Based on the theoretical threshold of 85 kg of cobalt element content in the bill of materials, the smart contract compares the actual extraction quantity of 92 kg reported by the recycler. When the limit is exceeded for 8 hours, activate the supplier performance deduction rule and send an alarm instruction to the supervision node in region N. The temperature and humidity data collected by the environmental sensor are bound to the disassembly batch, and a credibility score associated with the time stamp is generated and written into the verification parameter library. When the score is lower than 70 points, trigger the data review process.

[0037] Furthermore, the blockchain-based supply chain carbon data trustworthy management method described in the present invention further includes: collecting the monitoring data of the warehouse temperature and humidity sensor within a preset time window, and generating an environmental parameter time series bound to the product recycling batch code;

[0038] Collecting the monitoring data of the warehouse temperature and humidity sensor within a preset time window is mainly based on the following technical purposes:

[0039] Achieve environmental compliance of the recycling process: Temperature and humidity are key parameters affecting the quality of material recycling. For example, too high humidity during the battery disassembly process may cause metal corrosion, and abnormal temperature may accelerate the decomposition of chemical substances, affecting the recovery rate. Through continuous monitoring within a preset time window (such as non-operation periods), it can be verified whether the recycling operation is carried out under compliant environmental conditions, avoiding material loss or pollution risks caused by environmental abnormalities.

[0040] Verify the credibility of recycling data: After binding the temperature and humidity data to the recycling batch code to generate a time series, combined with an anomaly detection model (such as an LSTM model) trained with historical data, abnormal energy consumption fluctuations during non-operation periods can be identified. For example, if the temperature and humidity sensor shows that the warehouse is airtight but the energy consumption suddenly increases during the early morning hours, it may indicate unauthorized equipment operation or data tampering behavior, thus triggering a decrease in the data credibility score and starting the review process.

[0041] Support full-chain traceability: The time series of environmental parameters provides a complete "environmental fingerprint" for each recycling batch. When the actual recycling volume deviates from the theoretical threshold (such as abnormal cobalt extraction), the environmental data of the corresponding batch can be traced back to determine whether temperature and humidity affect the metallurgical process efficiency or whether there is human intervention (such as false reporting of recycling volume), providing an objective basis for auditing.

[0042] Optimize energy consumption analysis: Temperature and humidity are directly related to the energy consumption of warehousing (such as the power consumption of air conditioning systems). By analyzing the correlation between environmental data and energy consumption within a preset time window, the carbon emission calculation model in the recycling link can be dynamically calibrated to avoid carbon accounting deviations caused by abnormal operation of environmental regulation equipment and improve the accuracy of supply chain carbon footprint tracking.

[0043] The temperature and humidity data of the temperature and humidity sensor are cross-chain bound with the recycling process data through the blockchain smart contract, forming an immutable "environment-energy consumption-recycling" multi-dimensional evidence chain. For example, when the actual recycling volume exceeds the threshold and the environmental data is abnormal, the smart contract automatically deducts the supplier performance score and sends an encrypted alarm to the regulatory node, realizing a fully automatic closed-loop management from environmental monitoring to audit decision-making.

[0044] The energy consumption anomaly detection model trained based on the historical recycling data of the environmental parameter time series is used to identify the energy consumption fluctuation characteristics during non-operating periods;

[0045] The energy consumption fluctuation characteristics are associated and matched with the timestamps of the recycling process data, and a credibility score is generated and written into the verification parameter library of the blockchain smart contract.

[0046] The energy consumption fluctuation analysis is realized through LoRa sensors deployed in the warehouse in Region O, and the environmental data during non-operating periods is collected every 5 minutes. The LSTM model trained based on historical recycling data identifies the abnormal energy consumption characteristics from 1:00 to 5:00 in the early morning, and generates an auxiliary verification factor when the fluctuation amplitude exceeds 30% of the baseline value. After this factor is matched with the timestamp of the recycling process data, the supplier credit rating parameters in the smart contract are updated.

[0047] Furthermore, in the blockchain-based supply chain carbon data trustworthy management method of the present invention, the step S103 includes: setting at least two consensus nodes with auditing functions in the blockchain network, and configuring the carbon emission auditing permission identifier in the node certificate;

[0048] When the data correction instruction involves the update of the industry benchmark threshold, extract the corrected data features to generate a hash value to be signed;

[0049] The consensus nodes are used to call the elliptic curve digital signature algorithm to jointly sign and verify the hash value, and anchor the signature result to the blockchain transaction block header.

[0050] The multi-party signature process configures two audit nodes in the blockchain network of Region P. The node certificates include the audit permission identifiers issued by the SM2 algorithm. When the industry benchmark threshold is updated from 0.55 kgCO2 / kWh to 0.51 kgCO2 / kWh, the corrected data features are extracted to generate the hash value to be signed. The audit nodes call the SM9 algorithm for joint signature verification, and the signature digest is written into the block header and synchronized to the shard chain in Region Q. The transaction confirmation time is compressed within 2 seconds.

[0051] Furthermore, the blockchain-based trusted management method for supply chain carbon data of the present invention further includes: calculating the percentage deviation rate of the carbon emissions of each node from the industry benchmark value based on the real-time carbon emission data of each link in the supply chain;

[0052] Through spatial clustering analysis of the percentage deviation rate of the carbon emissions of each node from the industry benchmark value by the geographic information system, identifying the characteristics of regional concentrated abnormal distribution;

[0053] Writing the production capacity adjustment suggestions including the regional identifier into the intelligent contract execution queue, and the intelligent contract execution queue triggers the update of the supplier evaluation parameters according to the first-in, first-out rule.

[0054] The production capacity adjustment strategy is implemented based on the geographic information system of Region R, and the percentage deviation rate of the carbon emissions of each node in the supply chain from the industry benchmark is calculated. Spatial clustering analysis identifies that 3 suppliers in Region S have a deviation of more than 15%, and generates production capacity compression suggestions to be written into the intelligent contract queue. The queue triggers the update of the evaluation parameters according to the FIFO rule, and performs the supplier grading adjustment operation every 24 hours.

[0055] Furthermore, for the blockchain-based trusted management method for supply chain carbon data of the present invention, the construction of the dynamic emission factor model includes:

[0056] Collecting the thermal efficiency value of production equipment, the energy consumption coefficient of transportation vehicles, and the conversion rate of recycled materials through the industrial Internet of Things interface to generate a standardized parameter data set;

[0057] Inputting the standardized parameter data set into the genetic algorithm optimization engine, and performing chromosome crossover and mutation iteration under the emission constraint conditions that meet the GHG Protocol Scope 1-3;

[0058] Through the blockchain consensus nodes, perform compliance voting verification on the generated candidate emission factor set, and select the parameter combination with a voting passing rate exceeding the preset threshold to update the coefficients of the dynamic emission factor model.

[0059] The construction of dynamic emission factors is implemented in Region T, and parameters such as the thermal efficiency value of 82% of stamping equipment, the energy consumption coefficient of 0.35 L / km for transport vehicles, and the conversion rate of 91% for recycled materials are collected through the OPC UA interface. The genetic algorithm optimization engine iteratively generates a candidate factor set under the constraints of GHG Protocol Scope 1-3. The blockchain consensus nodes vote on the compliance of the candidate set, and the parameter combinations with a voting pass rate exceeding 75% update the coefficients of the dynamic emission factor model. The data during the optimization process is written into the quantum-resistant on-chain log.

[0060] During the construction of the dynamic emission factor model, multi-source parameters such as the thermal efficiency value of 83% of stamping equipment in Region A, the energy consumption coefficient of 0.38 L / km for transport trucks in Region B, and the conversion rate of 89% for the material conversion of the recycling production line in Region C are collected in real time through the industrial Internet of Things interface, and a standardized data set is generated and input into the genetic algorithm optimization engine. The engine simulates the biological evolution mechanism, encodes the candidate solutions of carbon emission factors into chromosome gene sequences, and iteratively generates a candidate factor set through the roulette wheel selection strategy and two-point crossover mutation operation under the combined constraint conditions of meeting GHG Protocol Scope 1 (direct emissions), Scope 2 (indirect emissions from purchased energy), and Scope 3 (other indirect emissions in the supply chain). In each round of the iterative process, the fitness function calculates the comprehensive score of the candidate factors in the energy efficiency verification scenario of transport vehicles and the thermal loss simulation scenario of production equipment, eliminates invalid solutions that deviate from the emission constraint boundary by more than 20%, and finally outputs the parameter combinations optimized through 50 generations of evolution and submits them to the blockchain consensus nodes for compliance voting verification.

[0061] Advantages of the present invention;

[0062] The present invention realizes the isolated storage and cross-chain synchronous verification of multi-regional data in the supply chain through blockchain sharding chain technology, effectively solving the problem of carbon emission accounting deviation caused by data islands in existing methods. The dynamic emission factor model combines equipment operation parameters to correct the calculation logic in real time, improving the accuracy of process improvement effect verification. The smart contract engine triggers anomaly detection based on industry benchmark thresholds and enhances the credibility of data correction through a multi-party signature confirmation mechanism. The edge computing nodes cooperate with the machine learning model to realize the correction of transport path deviation and the intelligent identification of energy consumption fluctuations in recycling processing. The genetic algorithm optimization engine generates a candidate emission factor set under the multi-scope constraints of GHG Protocol, and dynamically updates the model parameters in combination with the blockchain consensus verification mechanism, optimizing the accuracy of supply chain carbon footprint tracking. The linkage between the spatial clustering analysis of the geographic information system and the smart contract execution queue realizes the rapid positioning of regional carbon emission anomalies and the precise implementation of production capacity adjustment strategies, forming a reliable carbon data management closed-loop covering the entire life cycle. Description of the Drawings

[0063] To more clearly illustrate the technical solution of the present invention, the following will briefly introduce the drawings required for the embodiments. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the drawings.

[0064] Figure 1 It is a flowchart of a blockchain-based trusted management method for supply chain carbon data provided by an embodiment of the present invention. Specific embodiments

[0065] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention in conjunction with specific embodiments and corresponding drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention. The following will describe in detail the technical solutions provided by each embodiment of the present invention with reference to the drawings. To better understand the objectives of the present invention, the following will describe the present invention in more detail.

[0066] In a first aspect, please refer to Figure 1 , the present invention provides a blockchain-based trusted management method for supply chain carbon data, including:

[0067] Step S101, obtaining data of all links in the supply chain. The data of all links in the supply chain includes supplier production data, manufacturing process data, logistics transportation data, and recycling and processing data. Among them, the supplier production data collects the energy consumption in real time through an edge computing gateway deployed on the supplier production equipment. The manufacturing process data includes energy equipment sensor data and process improvement records in the factory. The logistics transportation data obtains the real-time working condition parameters of the transport vehicle through a vehicle-mounted terminal. The recycling and processing data includes the product disassembly material recovery rate;

[0068] The edge computing gateway establishes a communication connection with the data acquisition terminal of the supplier production equipment through an industrial Internet of Things protocol, and collects the consumption data of energy media such as electricity, gas, and steam in real time. The gateway is built with a data preprocessing module, which performs range conversion and unit unification on the original signal, eliminates outliers caused by sensor drift or communication interference, and generates a structured energy consumption data stream including device identifiers and timestamps. The energy consumption data stream is uploaded to the edge node of the blockchain network through an encrypted transmission channel, forming a data input source that is spatiotemporally aligned with the manufacturing process data.

[0069] Integrate the output of the sensor network of energy equipment in the factory during the manufacturing process, including the power curve of the air compressor, the thermal efficiency parameters of the boiler, and the time-series record of the circulation volume of the cooling system. The process improvement records are extracted through the enterprise resource planning system interface, and the record content covers the time nodes of equipment transformation, the adjustment amount of technical parameters, and the summary of the energy efficiency test report. The data fusion module performs versioned association on the sensor data and the process records to construct a dynamic data set reflecting the relationship between equipment status and process evolution, which serves as the input condition for the subsequent training of the carbon emission factor model.

[0070] The logistics transportation data collects the real-time operating conditions of the transportation vehicle through the multi-modal sensors of the vehicle-mounted terminal, including the engine speed, instantaneous fuel consumption, and load change curve. The vehicle-mounted terminal uploads data in a hybrid mode of event-triggered and periodic polling. When detecting events such as sudden acceleration, overloading, or route deviation, it preferentially transmits high-priority data packets. The transportation data stream collaborates with the roadside unit through the near-field communication protocol to complement the data missing segments caused by insufficient network coverage and form continuous spatio-temporal trajectory information.

[0071] The recycling and processing data obtains the classification recovery rate of metal components during the product disassembly process through the radio frequency identification system and weighing device deployed on the disassembly production line. The disassembly batch code and the material recovery rate data are associated with the product life cycle assessment report through the blockchain oracle interface to generate a traceable data unit including the material flow and the recovered carbon footprint. The recycling data and the carbon emission records in the production and transportation links are aligned through a unified time benchmark to form a carbon data topology network covering all links of the supply chain.

[0072] Step S102, input the data of all links of the supply chain into the blockchain network for trust processing, generate a data summary including spatio-temporal features through the data standardization module, and write the data summary into the blockchain distributed ledger. Among them, the following processing is performed on the supplier production data: call the regional carbon emission factors in the industry benchmark database for data deviation verification, and generate an on-chain record including a verification identifier; construct a dynamic emission factor model for the manufacturing process data in combination with equipment operation parameters, and verify the logical relevance between the process improvement record and the energy consumption through the dynamic emission factor model;

[0073] The blockchain network performs format parsing and semantic alignment on the input all-link data through the data standardization module, and uniformly converts the energy consumption unit in the supplier production data into the standard carbon emission equivalent. The spatio-temporal feature extraction engine generates a structured data summary including the regional code and time window identifier based on the geographical location coordinates of the equipment and the data collection timestamp. The data summary generates a unique summary identifier through the collision-resistant hash function and is written into the distributed ledger of the blockchain after being bound to the meta-information of the original data.

[0074] The trust processing of supplier production data calls the API interface of the industry benchmark database to dynamically match the grid carbon emission factor and production process type characteristics in the region where the supplier is located. The data deviation verification module uses a sliding window statistical model to calculate the dynamic deviation rate of real-time energy consumption data from the benchmark value. When the deviation rate exceeds the preset threshold, an on-chain record including the anomaly level and verification timestamp is generated. The encoding rule of the verification identifier integrates the blockchain transaction hash and the supplier identity information to form traceable verification status metadata.

[0075] The processing flow of manufacturing process data extracts key parameters such as the air compressor load rate, boiler thermal efficiency, and cooling system circulation volume through the device operation parameter parsing engine. The dynamic emission factor model constructs an association matrix of process improvement and energy consumption change based on parameter time-series data, and uses a logistic regression algorithm trained with historical data to verify the influence trend of process adjustment on carbon emission intensity. The output layer of the model generates an emission factor correction value including the version number and confidence interval, which is updated to the on-chain accounting rule library through a smart contract, triggering the subsequent audit process to verify the compliance of the correction logic.

[0076] The blockchain distributed ledger synchronizes data digests and verification records to multiple nodes through a consensus mechanism. The ledger structure design adopts a sharding storage strategy, and writes supplier production data, manufacturing process model parameters, and logistics transportation records into independent data channels respectively. The block generation interval of each channel sets different frequencies according to the data type. The production data channel adopts a block generation strategy at the second level, and the model parameter update channel adopts a minute-level batch confirmation mechanism to balance the requirements of data real-time and system throughput. The on-chain records after writing are opened to the regulatory node through a cross-chain query interface to support the traceability verification of the carbon emission data throughout the life cycle.

[0077] Step S103, input the on-chain record into the smart contract execution engine to make an audit decision based on the preset carbon emission compliance rules. When it is detected that the verification identifier corresponding to the energy consumption deviates from the industry benchmark threshold, trigger the generation process of the data correction instruction and execute multi-party signature confirmation through the blockchain consensus node;

[0078] The smart contract execution engine loads the preset carbon emission compliance rule library, and the rule library adopts a hierarchical logical structure, including industry emission standards, regional regulatory policies, and enterprise-defined constraint conditions. The audit decision module performs pattern recognition on the verification identifier of the on-chain record based on the matching priority of the rule library, and uses a sliding window mechanism to analyze the time-series deviation characteristics of the energy consumption. When the number of consecutive deviations exceeds the preset threshold, trigger the anomaly event classifier to generate the metadata of the data correction instruction.

[0079] The data correction instruction generation process calls the transaction construction interface of the blockchain network, encapsulating the corrected target data field, baseline threshold version number, and deviation magnitude into a structured transaction request. The consensus nodes verify the transaction legality based on the preset audit permission level, and perform a distributed signature operation on the hash digest of the correction instruction using the threshold signature mechanism. The signature process introduces a time constraint protocol, requiring at least two nodes with different regulatory permissions to complete collaborative signatures within a preset time window. The signature shards generate a complete composite digital signature through an aggregation algorithm.

[0080] The correction instruction that passes the signature verification updates the on-chain data record through the smart contract state machine. The update operations include incrementing the version number, writing the correction timestamp, and saving the snapshot of the previous data state. The updated data hash value is anchored to the new block header through the Merkle tree structure. The block header synchronously records the certificate identifiers and timestamp sequences of the signature participating nodes, forming a traceable correction operation chain. After the audit log generation module captures the state change event, it generates log entries including the comparison of data before and after correction, the list of signature nodes, and the trigger rules, and writes them into the immutable storage layer of the blockchain for subsequent regulatory audit calls.

[0081] Step S104, generate a supply chain optimization instruction according to the result of the audit decision, and feedback the supply chain optimization instruction to the management-side execution device, where the supply chain optimization instruction includes supplier evaluation parameters and transportation route planning strategies dynamically adjusted based on the anomaly detection result, forming a low-carbon supply chain management closed-loop system that runs through data collection, verification, and decision-making.

[0082] The supply chain optimization instruction generation module dynamically adjusts the weight coefficients of the supplier evaluation parameters using the weighted priority algorithm based on the anomaly event distribution characteristics in the audit decision result. The evaluation parameter set integrates historical performance records, carbon emission deviation rates, and rectification response timeliness indicators, and calculates the grading adjustment range of the supplier credit rating through a rule engine. The transportation route planning strategy calls the digital twin system to model the multimodal transport network, combines real-time traffic flow data with the regional carbon emission intensity layer, and generates an optimization plan set including the carbon emission simulation values of alternative routes.

[0083] The management-side execution device receives the optimization instruction through the cross-chain communication protocol of the blockchain network. The instruction data packet is encapsulated into a structured transaction including the execution time window, target parameter values, and constraint conditions. The supplier evaluation parameter update task is written into the supplier management database through the enterprise resource planning system interface, triggering the automatic adjustment of the supplier production schedule and energy procurement strategy. The transportation route optimization instruction is distributed to the route planning engine of the logistics scheduling platform, forming a feedback control loop with the real-time navigation data of the vehicle terminal to dynamically correct the driving route of the transport vehicle.

[0084] The closed-loop management system continuously obtains information on the status of suppliers' equipment, logistics trajectories, and recycling and processing progress through the data acquisition layer, and inputs it into the blockchain network to execute a new round of data standardization and trustworthiness processing procedures. The verification layer performs secondary verification on the processed data based on the updated dynamic emission factor model, and the decision-making layer triggers the iterative generation of optimization instructions based on the real-time verification results. The status change events of data in each link are synchronized to the audit log module through the event bus, forming a traceability chain covering the entire cycle of instruction execution, and realizing the self-adaptive optimization of the carbon data management system.

[0085] In step S101, the energy consumption of suppliers' production equipment is collected in real time by deploying edge computing gateways, and multi-source heterogeneous data is synchronously obtained in combination with industrial Internet of Things protocols. The edge computing gateways integrate device-level data preprocessing functions, perform preliminary cleaning and format conversion on the original energy consumption data, and eliminate noise data caused by sensor sampling errors or communication interference. Logistics transportation data collects real-time operating parameters of transportation vehicles through in-vehicle terminals, including engine speed, load changes, and driving trajectory coordinates, and is transmitted to the cloud data center through lightweight communication protocols. Radio frequency identification devices and material weighing systems are deployed in the recycling and processing link to obtain the recovery rate data of each metal component during the product disassembly process, forming a multi-modal data acquisition network covering the entire life cycle of the supply chain.

[0086] In step S102, the blockchain network extracts spatio-temporal features from multi-source heterogeneous data through the data standardization module, and generates a structured data summary including geographical coordinates and timestamps. The production data of suppliers calls the regional carbon emission factors in the industry benchmark database, and uses the dynamic threshold comparison algorithm to perform data deviation verification. The verification results are written into the blockchain distributed ledger in the form of encrypted identifiers. The manufacturing process data constructs a dynamic emission factor model based on the two-way mapping of equipment operation parameters and process improvement records, and this model verifies the correlation between process parameter adjustments and energy consumption changes through the on-chain verification mechanism, generating a carbon emission accounting rule set with version identifiers. Logistics transportation data generates a theoretical optimal carbon emission benchmark value through a path planning algorithm, and forms a cross-verification mechanism with the actual operating condition data collected by in-vehicle sensors.

[0087] In step S103, the smart contract execution engine loads the preset carbon emission compliance rule library, performs pattern recognition and anomaly detection on the multi-dimensional data recorded on the chain. When the verification identifier deviates from the industry benchmark threshold, it triggers a data correction process based on multi-party collaboration: the consensus nodes jointly verify the correction instructions based on the elliptic curve digital signature algorithm, and the verified correction data generates a new hash feature value and updates the on-chain status. The audit decision module synchronously analyzes the version iteration records of the dynamic emission factor model in the manufacturing process, evaluates the impact weight of process improvement measures on the carbon emission accounting rules, and generates an audit report covering upstream and downstream nodes of the supply chain.

[0088] In step S104, based on the distribution characteristics of abnormal events in the audit report, the supply chain optimization instruction dynamically adjusts the weight coefficients of supplier evaluation parameters using a machine learning algorithm. The logistics path planning strategy integrates the spatial topological data of the multimodal transport carbon emission tracking map, combines the alternative path simulation results generated by the digital twin system, and generates an optimization plan that takes into account both transportation timeliness and carbon emission reduction goals. After the management end executes the device receiving optimization instruction, it distributes the parameter adjustment instruction to the supplier production control system and the logistics scheduling platform through the cross-chain communication protocol, forming a closed-loop control system from data collection, on-chain verification to execution feedback. The real-time operation status data of each link is continuously input into the blockchain network, driving the periodic iterative update of the dynamic emission factor model and the compliance rule library, and realizing the adaptive optimization of the supply chain carbon management system.

[0089] Specifically, for the blockchain-based trusted management method of supply chain carbon data described in the present invention, step S102 further includes:

[0090] Call the raw material procurement database of a third-party certification agency through the blockchain oracle interface, extract the hash digest of the original procurement voucher, and generate a hash feature value including a timestamp;

[0091] Input the hash feature value and the bill of materials data in the life cycle assessment report into the zero-knowledge proof verification module to verify the relevance and consistency of the procurement voucher and the carbon emission inventory;

[0092] Generate a blockchain data block including a trusted identifier according to the verification result, and the trusted identifier is used to update the integrity evaluation status of supplier production data.

[0093] The blockchain oracle interface establishes a communication link with the third-party raw material procurement database through a preset authentication protocol, and uses asymmetric encryption technology to protect the transmission channel end-to-end. The original procurement voucher is parsed by the data parsing module to extract key fields, including the supplier code, material batch, and procurement quantity, generates a unique hash digest through a collision-resistant hash function, and embeds the timestamp information provided by the trusted time source into the digest encoding sequence to form an immutable hash feature value. This feature value is transmitted to the zero-knowledge proof verification module through the cross-chain routing protocol as an input parameter for the verification process.

[0094] The zero - knowledge proof verification module constructs a verification relationship based on the preset circuit logic, and logically maps the hash eigenvalue to the bill of materials data in the life cycle assessment report. The verification process simulates the correlation constraints between the procurement voucher and the carbon emission inventory by constructing an arithmetic circuit, and generates a verifiable zero - knowledge proof result without revealing the original data. When there is a logical conflict between the raw material carbon footprint data marked in the bill of materials and the procurement record corresponding to the hash eigenvalue, the verification module outputs a verification report including an exception identifier, triggering the on - chain data correction process.

[0095] The trusted identifier generation module writes metadata including the verification status code into the blockchain data block according to the zero - knowledge proof verification result. The encoding rule of the trusted identifier integrates the supplier identity information, verification timestamp and the hash value of the associated data block, and updates the integrity assessment status of the supplier production data through a smart contract. The integrity assessment status is recorded in the on - chain node in the form of a dynamic weight value, which is used as a key input parameter for the supplier credit rating in the subsequent audit decision process. The blockchain network broadcasts the updated assessment status to all nodes through the synchronization mechanism of the distributed ledger, supporting all participants in the supply chain to retrieve the verification records in real time.

[0096] Specifically, for the blockchain - based trusted management method of supply chain carbon data described in the present invention, step S102 includes

[0097] Access the waste disposal blockchain of the regulatory platform through a cross - chain relay node to obtain the digitally signed electronic consignment note data;

[0098] Extract the time - series consumption data of the chemical inventory management system at the production end, and calculate the dynamic deviation rate between the time - series consumption data and the disposal volume of the electronic consignment note;

[0099] When the dynamic deviation rate exceeds the preset threshold, trigger the smart contract to lock the on - chain status of the carbon emission calculation result for the corresponding production batch.

[0100] The cross - chain relay node establishes a communication connection with the waste disposal blockchain of the government regulatory platform through a preset cross - chain protocol, and verifies the legality of the regulatory chain nodes using a distributed identity authentication mechanism. After the electronic consignment note data is parsed by the digital signature verification module, structured fields including waste types, disposal volume and timestamp are extracted, and format conversion and semantic alignment between heterogeneous blockchains are realized through the data cache layer of the relay node. The electronic consignment note data of the regulatory chain is synchronized to the verification node of the enterprise production chain in a light - node mode, forming a verifiable reference relationship for cross - chain data.

[0101] The production-side chemical inventory management system collects time-series consumption data through the industrial Internet of Things interface, including raw material input, waste generation rate, and storage container status parameters. The time-series data processing module performs sliding window statistics on the consumption data, calculates the average consumption rate within the unit time window, and normalizes it with the disposal volume recorded in the electronic form. The dynamic deviation rate calculation engine generates quantitative indicators that reflect the degree of matching between consumption and disposal based on the preset sliding window step size and weight distribution strategy. The time series similarity algorithm is introduced into the indicator calculation process to eliminate the impact of equipment start-stop fluctuations on data comparison.

[0102] When the smart contract detects that the dynamic deviation rate exceeds the preset threshold, it calls the carbon emission calculation contract state machine stored on the chain to perform a read-only lock operation on the carbon emission data corresponding to the abnormal production batch. The locked state is updated to all nodes through the blockchain consensus mechanism, limiting the subsequent modification permissions of the batch data, and generating an audit trail record including the lock reason and timestamp. After the state lock is triggered, the off-chain collaborative verification module starts the multi-source data review process, integrates the production equipment operation log and waste transportation trajectory data, and generates a supplementary verification report for subsequent manual review and decision-making. The release of the locked state requires the joint signature verification of the review results by the multi-party consensus node to reactivate the state update function of the smart contract.

[0103] Specifically, the blockchain-based supply chain carbon data trusted management method of the present invention, step S102 includes:

[0104] Based on the starting or ending coordinates of the transport task and the vehicle load parameters, the path planning algorithm is called to generate the benchmark carbon emission value of the theoretical optimal transport path;

[0105] Obtaining real-time operating data of the actual transport trajectory through on-board sensors, and calculating the incremental difference between the actual carbon emissions and the benchmark carbon emissions value;

[0106] When the incremental difference exceeds the path deviation threshold preset in the smart contract, an abnormal transportation event record is created in the blockchain ledger, and a suspension instruction of the on-chain freight settlement contract is triggered.

[0107] The path planning engine generates the topological structure of the theoretically optimal transportation path based on the geographic fence parameters of the starting and ending coordinates of the transportation task, combined with the nonlinear relationship model of the impact of vehicle load on energy consumption. The path planning algorithm adopts a multi-objective optimization strategy, integrates road network topology data and historical transportation energy consumption characteristics, and constructs a decision model with carbon emission minimization as the core constraint condition, outputting a benchmark carbon emission value including the path node sequence and expected energy consumption intensity. The regional carbon emission factor weight parameter is embedded in the process of generating the benchmark carbon emission value to achieve dynamic adaptation of transportation mode and geographical characteristics.

[0108] On-vehicle sensors collect engine speed, fuel consumption rate, and GPS positioning data in real time, and the edge computing node performs noise filtering and time alignment processing on the original working condition data. The real-time working condition data analysis module inputs the cleaned data stream into the carbon emission calculation model, combines the load change curve and the instantaneous vehicle speed parameters, and generates the actual carbon emission time series indexed by the time stamp. The incremental difference calculation engine, based on the sliding window mechanism, dynamically matches the discrete points of the reference carbon emission value and the actual value, and uses the normalization algorithm to eliminate the interference of the path segmentation error on the deviation calculation, generating a quantitative index reflecting the deviation degree of carbon emissions during the transportation process.

[0109] When the intelligent contract monitors that the incremental difference exceeds the path deviation threshold, it calls the transaction generation interface of the blockchain ledger to create an abnormal transportation event record. The event record includes the transportation task identifier, the deviation magnitude, and the trigger time stamp, and is anchored to the immutable storage layer of the blockchain through the Merkle tree structure. The state machine of the freight settlement contract on the chain synchronously receives the abnormal event trigger signal, executes the conditional judgment logic in the contract code, and sets the settlement instruction queue to the pause state. During the suspension of execution, the blockchain network starts a multi-party verification process, integrates the historical path optimization data and the vehicle maintenance records, and generates an abnormal event review report for subsequent manual intervention or automated decision callback.

[0110] Specifically, the blockchain-based trusted management method for supply chain carbon data described in the present invention further includes: parsing the energy consumption log data of port handling equipment through an edge computing node, and extracting the ship speed and cargo volume parameters in the ship AIS trajectory data;

[0111] Synchronously collecting the GPS trajectory data and load change parameters of road transport vehicles, the freight formation information and energy consumption monitoring data of railway transport;

[0112] Constructing a carbon emission calculation model for the sea transportation segment based on the ship speed and cargo volume parameters, and integrating the real-time load, path optimization parameters of road transport, and the energy consumption coefficient of railway transport to generate a carbon emission integration model covering multimodal transport;

[0113] Extracting the carbon emission intensity characteristics of sea, road, and railway transport, and associating and mapping them through a preset coding rule with the transport identifier;

[0114] Generating a carbon emission feature vector including ship identifiers, road transport batch codes, and railway freight waybills to realize the binding of carbon footprint identifiers across transport modes.

[0115] Cross-chain hash binding the carbon emission feature vector with the blockchain data of road, railway, and sea transportation segments, and generating a carbon footprint tracking map covering multiple transport modes of road, railway, and sea through an intelligent contract.

[0116] Edge computing nodes access the energy consumption monitoring system of port handling equipment through a preset industrial communication protocol, and analyze the associated data of power consumption and operation time in the equipment operation logs. The Automatic Identification System (AIS) trajectory data extracts speed, cargo volume, and navigation status parameters through a data cleaning module, and uses a time series alignment algorithm to convert discrete AIS message data into a continuous navigation feature curve. The parsed ship energy consumption data is matched with the handling equipment operation logs through a spatio-temporal association model to generate input parameters for calculating carbon emissions in the sea transportation section, including the port operation stage and the sea transportation stage.

[0117] The GPS trajectory data of road transport vehicles is uploaded through in-vehicle terminals in an event-driven mode. Edge computing nodes construct a load-path association matrix based on the real-time weight change curve collected by load sensors. The freight formation information of railway transportation is obtained through the open interface of the railway dispatching system, and the train traction energy consumption monitoring data and the carriage formation topology relationship are integrated to generate an energy consumption feature set for the railway section based on the traction force distribution model. The synchronous acquisition of multi-source heterogeneous data is realized by a distributed message queue, and a cross-transport mode data association index is established through data timestamps and task identifiers.

[0118] The carbon emission calculation model for the sea transportation section is based on the ship navigation resistance formula and the non-linear relationship between cargo volume and speed, and constructs a carbon emission intensity curve per unit of cargo turnover. The road transport model introduces the influence coefficient of real-time load on vehicle rolling resistance, and combines the road slope and traffic flow parameters output by the path planning engine to dynamically correct the carbon emission calculation logic. The railway transport model generates a carbon emission allocation rule based on the carriage formation unit according to the power distribution ratio and traction energy consumption characteristics of the freight formation. The multi-modal carbon emission fusion model integrates the characteristics of each transport section through a weight allocation strategy, and uses a sliding window mechanism to dynamically adjust the contribution weight of the road, railway, and sea models.

[0119] The carbon emission feature vector generation module uses the ship IMO number, road transport batch code, and railway freight waybill number as the primary key identifiers, and embeds the carbon emission intensity values, timestamps, and geographical fence coordinates of each transport section. The data structure design of the feature vector supports the cross-chain query interface specification, and realizes the data semantic interoperability of different blockchain networks through a standardized coding format. The cross-chain hash binding protocol decomposes the feature vector into road, railway, and sea sub-vectors, and performs binding operations with the blockchain transaction hash values of the corresponding transport chains respectively to generate an indivisible combined hash digest.

[0120] The smart contract invokes the multimodal transportation tracking graph generation algorithm based on the combined hash digest. The graph nodes represent the carbon emission data units at each stage of the transportation task, and the edge relationship characterizes the carbon footprint connection logic when the transportation mode is switched. During the graph generation process, a time window constraint mechanism is introduced to dynamically aggregate the multi-transport segment data within a preset period to form a traceable spatio-temporal evolution structure of the carbon footprint. The blockchain network conducts a consensus verification on the integrity of the graph data through cross-chain verification nodes, and the verified graph metadata is written into the distributed ledger to support the supply chain participants to query the full-link carbon footprint topological relationship according to their permissions.

[0121] Specifically, for the blockchain-based trusted management method of supply chain carbon data described in the present invention, the step S103 includes:

[0122] Based on the metal element content parameters marked in the product design bill of materials, calculate the theoretical recovery threshold according to the metallurgical recovery rate formula;

[0123] Input the actual extraction amount reported by the recycler and the theoretical recovery threshold into the blockchain smart contract for deviation comparison;

[0124] When it is detected that the actual extraction amount exceeds the theoretical threshold and the duration exceeds the preset period, activate the supplier performance deduction rule module in the smart contract and send an alarm instruction including the abnormal batch identifier to the blockchain node with supervision authority.

[0125] The product design bill of materials extracts the metal element content parameters through the structured data analysis module, including the element type, mass percentage, and material batch identifier. The metallurgical recovery rate formula constructs a calculation model including melting efficiency, slag loss rate, and purity correction factor based on the metal smelting process characteristics and the recycler production line equipment parameters, and outputs the theoretical recovery threshold. During the generation process of the theoretical threshold, a feature weight coefficient trained with historical recovery data is introduced to dynamically adapt to the differences in the physical and chemical properties of different metal components.

[0126] The blockchain smart contract deploys a deviation comparison logic module, receives the data stream of the actual extraction amount reported by the recycler, and converts the unstructured report into a numerical sequence with a unified measurement unit through the data standardization interface. The deviation comparison engine uses a sliding window mechanism to calculate the dynamic deviation rate of the actual extraction amount and the theoretical threshold. The window step size is set to be synchronized with the recycling operation cycle to eliminate the interference of short-term fluctuations within the batch on the abnormal determination. The comparison result is written into the on-chain state database in the form of a timestamp index to support cross-batch data backtracking analysis.

[0127] When it is detected that the actual extraction volume exceeds the theoretical threshold continuously and the duration reaches the preset period, the smart contract calls the supplier performance deduction rule engine. Based on the preset deduction gradient table and the cumulative frequency of abnormal events, the rule engine generates a dynamically adjusted performance deduction value and updates it to the supplier credit rating chain data set. The alarm instruction generation module simultaneously creates a structured alarm message including abnormal batch code, overscalar level and time interval, and transmits it to the supervision node through the permission filtering channel of the blockchain node. After receiving the alarm, the supervision node activates the off-chain review process, retrieves the temperature and humidity sensor time series data of the corresponding batch and the disassembly operation video for evidence, and generates a composite verification report and writes it into the distributed file storage system.

[0128] Specifically, the blockchain-based supply chain carbon data trusted management method of the present invention further includes: collecting monitoring data of warehouse temperature and humidity sensors within a preset time window to generate an environmental parameter time series sequence bound to the product recycling batch code;

[0129] The energy consumption anomaly detection model trained based on the environmental parameter time series data of historical recycling data is used to identify the energy consumption fluctuation characteristics during non-operating periods;

[0130] The energy consumption fluctuation characteristics are associated and matched with the timestamp of the recycling processing data to generate a credibility score that is written into the verification parameter library of the blockchain smart contract.

[0131] The warehouse temperature and humidity sensors establish a communication connection with the edge computing node through the low-power wide area network protocol (LPWAN), and collect environmental parameters at a fixed sampling interval within the preset time window. The sensor node deployment strategy is based on the thermodynamic characteristics of the recycling operation area, and uses a spatial grid distribution method to cover the material storage area and the disassembly operation area to generate a multi-dimensional monitoring data stream bound to the recycling batch code. The time series generation module aligns the timestamps of the original data and removes outliers, constructs a feature vector including the mean, extreme value and change rate of temperature and humidity, and transmits it to the blockchain data preprocessing layer through an encrypted channel.

[0132] The energy consumption anomaly detection model trained based on historical recycling data uses time series segmentation technology to define non-operating periods as time windows when equipment is shut down and personnel activities are stationary. After the model input layer receives the time series of environmental parameters, it extracts the implicit correlation features between temperature and humidity fluctuations and basic energy consumption through the long short-term memory network (LSTM) to identify the characteristic patterns of abnormal energy consumption events, such as abnormal startup of the air conditioning system or unauthorized operation of ventilation equipment during non-operating periods. The model output layer generates a detection report including the anomaly type, confidence level, and start time, which is converted into a structured data format recognizable by the blockchain through the feature encoder.

[0133] The association matching module for fluctuation characteristics and recycling process data uses the dynamic time warping algorithm (DTW) to compare the timestamps of abnormal energy consumption events with the time intervals of disassembly operation records. A sliding window constraint mechanism is introduced during the matching process to eliminate the timing misalignment problems caused by equipment startup and shutdown delays or sensor clock deviations. The credibility score generation engine calculates the score value using a multi-dimensional weighting strategy based on the type weight, duration, and historical occurrence frequency of abnormal events. The score result is written into the on-chain database through the verification parameter update interface of the smart contract. The blockchain network broadcasts the change status of the verification parameter library to the entire chain through a cross-node synchronization mechanism, supporting subsequent audit processes to call the score data to perform supplier credit assessment and carbon emission accounting correction.

[0134] Specifically, for the blockchain-based trusted management method of supply chain carbon data described in the present invention, the step S103 includes:

[0135] Set at least two consensus nodes with auditing functions in the blockchain network, and configure the node certificates to include carbon emission audit permission identifiers;

[0136] When the data correction instruction involves the update of the industry benchmark threshold, extract the corrected data features to generate a hash value to be signed;

[0137] Call the elliptic curve digital signature algorithm through the consensus nodes to jointly verify the hash value, and anchor the signature result to the blockchain transaction block header.

[0138] The blockchain network deploys consensus nodes with auditing functions through a preset node access mechanism. The node certificates embed carbon emission audit permission identifiers in the X.509 standard format. The permission identifiers include verifiable audit scope and operation type constraint parameters. The geographical distribution strategy of the consensus nodes is designed based on the regulatory levels of supply chain participants, forming a decentralized audit network topology. The communication links between nodes are established using a national cryptography algorithm encrypted transmission protocol.

[0139] When the data correction instruction triggers the update of the industry benchmark threshold, the corrected data feature extraction module performs semantic parsing on the changed carbon emission factors, equipment parameters, or path optimization rules, generating a structured data packet including a version number and a change summary. The hash value generation engine for the data to be signed performs a digest operation on the structured data packet based on a quantum-resistant hash algorithm. The previous block hash is introduced as a random salt value during the operation process to generate a unique hash digest with chain association characteristics. The hash digest is pushed to the signature task queue of the audit node through an event-driven interface.

[0140] The consensus node invokes the key management module of the elliptic curve digital signature algorithm. After verifying the legality of the audit operation based on the permission identifier, it performs a distributed signature operation on the hash digest using the threshold signature mechanism. The joint signature verification process synchronizes the signature shards of each node through a preset multi-party collaboration protocol and aggregates them to generate a complete composite digital signature. The signature result is written into the transaction metadata field through the blockchain transaction constructor, and the transaction hash is anchored to the block header of the new block based on the Merkle tree structure. The historical signature digest of the previous audit operation is synchronously recorded in the block header to form an audit traceability chain data structure.

[0141] After being verified by the consensus algorithm, the transaction block with the anchored transaction is broadcast to all network nodes, and a two-way reference relationship is formed between the audit traceability chain data and the carbon emission accounting records of the main chain. The intelligent contract state machine updates the audit log based on the signature verification status in the block header. The log entries include the audit node identifier, signature timestamp, and associated data version number, supporting subsequent regulators to trace the full life cycle operation records of threshold changes through the cross-chain query interface.

[0142] Specifically, the blockchain-based trusted management method for supply chain carbon data described in the present invention further includes: calculating the percentage deviation rate of the carbon emissions of each node from the industry benchmark value based on the real-time carbon emission data of each link in the supply chain;

[0143] Performing spatial clustering analysis on the percentage deviation rate of the carbon emissions of each node from the industry benchmark value through a geographic information system to identify regional centralized abnormal distribution characteristics;

[0144] Writing the production capacity adjustment suggestion including the regional identifier into the intelligent contract execution queue, and the intelligent contract execution queue triggers the update of the supplier evaluation parameters according to the first-in, first-out rule.

[0145] The real-time carbon emission data of each link in the supply chain is normalized through a distributed data middleware, integrating multi-source heterogeneous data streams from production equipment sensors, logistics vehicle terminals, and recycling and processing systems. The percentage deviation rate calculation engine extracts the benchmark value matching the node's geographical location, production scale, and process type based on the dynamic update mechanism of the industry benchmark database, and uses a sliding window statistical model to calculate the carbon emission deviation rate within a continuous time window. The equipment operation status correction factor is introduced in the deviation rate calculation process to eliminate the influence of temporary shutdowns or load fluctuations on the statistical results.

[0146] The geographic information system performs spatial encoding conversion on the deviation rate dataset, maps the node geographical location information to the longitude and latitude coordinate grid, and identifies abnormal areas with spatial aggregation characteristics through the density clustering algorithm. The clustering analysis module generates a heat map reflecting the regional carbon emission deviation distribution based on the preset neighborhood radius and minimum sample number threshold, and marks the grid cells with heat values exceeding the critical threshold as abnormal clusters. The feature extraction engine for abnormal areas combines the supply chain topological relationship, analyzes the production process relevance and logistics path overlap degree of the nodes within the cluster, and generates a diagnostic report including the regional code, abnormal type, and influence range.

[0147] The production capacity adjustment suggestion generation module defines adjustment strategies according to the abnormal types in the diagnostic report. For example, it gives suggestions on optimizing the production rhythm or upgrading energy equipment in high deviation rate areas. The suggestion data is encapsulated into a standardized instruction format, with the regional identifier and priority weight attached, and is written into the execution queue through the smart contract interface. The queue management engine adopts the first-in-first-out scheduling strategy, sorts and triggers the supplier evaluation parameter update task according to the instruction priority and timestamp. The parameter update process calls the blockchain consensus mechanism to verify the compliance of the instructions, and the updated evaluation parameters are synchronized to the supplier production control system and the supervision platform through the cross-chain protocol, forming a closed-loop management mechanism for regional carbon emission anomalies.

[0148] Second, for the blockchain-based supply chain carbon data trusted management method of the present invention, the construction of the dynamic emission factor model includes:

[0149] Collect the thermal efficiency value of production equipment, the energy consumption coefficient of transportation vehicles, and the conversion rate of recycled materials through the industrial Internet of Things interface to generate a standardized parameter dataset;

[0150] Input the standardized parameter dataset into the genetic algorithm optimization engine, and perform chromosome crossover and mutation iterations under the emission constraint conditions that meet the GHG Protocol Scope 1-3;

[0151] Conduct compliance voting verification on the generated candidate emission factor set through the blockchain consensus node, and select the parameter combination with a voting passing rate exceeding the preset threshold to update the dynamic emission factor model coefficient.

[0152] The industrial Internet of Things interface establishes a data channel with the production equipment control system through the OPC UA protocol, and real-time collects the thermal efficiency value, equipment load rate, and start-stop status parameters. The energy consumption coefficient of the transport vehicle extracts the engine operating condition data stream through the in-vehicle CAN bus parsing module, and generates the dynamic energy consumption characteristics per unit of transport volume in combination with the load sensor and GPS positioning information. The material conversion rate in the recycling process collects the metal component content data through the weighing system and the spectral analyzer, and forms a standardized parameter sequence after removing noise through the data cleaning module. The multi-source heterogeneous data is aligned by timestamp and encoded by spatial grid, and integrated into a structured data set including equipment identifiers, transport task codes, and recycling batch numbers.

[0153] The genetic algorithm optimization engine encodes the standardized parameters into a chromosome gene sequence, and each gene locus corresponds to the optimization weight of a specific parameter type (such as thermal efficiency, energy consumption coefficient). The fitness function design integrates the composite constraint conditions of GHG Protocol Scope 1-3. The Scope 1 constraint focuses on the matching degree between the direct emission parameters and the equipment operating status. The Scope 2 constraint introduces the regional power grid carbon emission factor correction rule. The Scope 3 constraint integrates the indirect emission correlation model of the transport path and the recycling process. The chromosome crossover and mutation operations adopt the two-point crossover and adaptive mutation rate strategy, and iteratively generate a candidate emission factor set within the solution space that satisfies the emission constraint boundary. Each round of iteration eliminates the invalid solutions that deviate from the constraint threshold by more than 20%.

[0154] The blockchain consensus nodes perform compliance voting verification on the candidate emission factor set, and the voting weight distribution strategy sets different permissions according to the node type (such as regulatory nodes, enterprise nodes). The candidate factor set is distributed to each node through the smart contract. The node performs compliance verification based on the local verification rule library (such as industry emission standards, regional regulatory policies), and generates a voting record including the verification summary. The voting result aggregation module counts the candidate parameter combinations with a passing rate exceeding the preset threshold (such as 75%), and triggers the model coefficient update instruction. The updated emission factor coefficient is synchronized to the carbon emission calculation nodes in all links of the supply chain through the cross-chain protocol, and the version number, effective time, and associated voting summary are recorded in the blockchain distributed ledger, forming a traceable model iteration history chain.

[0155] The specific implementation manner of the present invention is as follows: In the production link of suppliers in Area A, the edge computing gateway deployed on stamping equipment collects power consumption data at a sampling frequency of once per second and transmits it to the data processing node through the industrial Internet of Things protocol. The logistics vehicles in Area B are equipped with GPS positioning terminals, record the transportation coordinates once per minute, and obtain the load parameters in real time in combination with in-vehicle sensors. The data is uploaded to the cloud through the MQTT protocol. The standardized processing module calls the power emission factor of 0.532 kgCO2 / kWh in Area A and the road transportation emission factor of 0.82 kgCO2 / km in Area B to convert the original energy consumption and transportation mileage into a unified carbon emission equivalent value. The anomaly detection model optimizes the feature weights through the genetic algorithm. When the carbon emission deviation between the energy consumption of the stamping process and the transportation mileage converted in the same batch exceeds 20%, it triggers the process of retrieving the historical data of the load sensor for verification.

[0156] In the metadata generation stage, the SHA-256 algorithm is used to perform a hash operation on the device timestamp and transportation trajectory coordinates to generate a 128-bit data identifier, and a structured tag including the supplier code and logistics batch is constructed. The blockchain sharding chain writes the data in Area A into the Hyperledger Fabric framework according to the supplier's registered location, and the data in Area B is stored in the FISCO BCOS framework. The block generation interval is set to 5 seconds, and the transaction confirmation time is controlled within 3 seconds. The smart contract engine calls the on-chain verification code every 15 minutes to compare the real-time carbon emission value with the industry benchmark parameters. When it is detected that the carbon emission of sea container transportation exceeds the benchmark value by 15%, it triggers an early warning mechanism to generate an alarm event including the container number.

[0157] After the edge computing node responds to the early warning, it retrieves the ship AIS trajectory data for cross-verification with the hash value stored in the blockchain, and preferentially uses the net weight of the goods recorded by the load sensor to correct the carbon emission record in the transportation link. When the daily data volume of the new supplier reaches the 10GB threshold, the system automatically creates an independent sharding chain based on the Quorum framework and synchronizes the Merkle Root hash values of each sharding chain through the cross-chain relay protocol. The dynamic monitoring module synchronizes and updates the benchmark parameters from the regulatory database every quarter. When the carbon emission increase caused by the change of the logistics path exceeds the threshold set by the smart contract, the digital twin system calls the Monte Carlo algorithm to generate three alternative path plans, and the screening conditions include that the transportation time deviation ≤ 48 hours and the carbon emission reduction ≥ 25%.

[0158] In the recycling process, the warehouse temperature and humidity sensors collect environmental data at a frequency of once every 5 minutes, generating a time series sequence bound to the recycling batch code. The machine learning model trained based on historical recycling data identifies the characteristics of energy consumption fluctuations during non-operating periods, associates the fluctuation amplitude with the timestamp, generates a credibility score, and writes it into the intelligent contract verification parameter library. When it is detected that the metal recycling volume exceeds the theoretical threshold and remains over-limit for 24 hours, the intelligent contract activates the supplier performance deduction rule and sends an encrypted alarm instruction to the regulatory node.

[0159] In the blockchain network, at least two consensus nodes with auditing functions are configured, and the node certificates include carbon emission auditing permission identifiers. When the industry benchmark threshold needs to be updated, the corrected data features are extracted to generate a hash value to be signed, and joint verification is performed through the elliptic curve digital signature algorithm. The signature result is anchored to the transaction block header. The geographic information system conducts spatial clustering analysis on the carbon emission deviation rates of each link in the supply chain. After identifying the concentrated anomalies in Region A, it generates production capacity adjustment suggestions and writes them into the intelligent contract execution queue, triggering the update of supplier evaluation parameters according to the first-in, first-out rule.

[0160] The above implementation method realizes the trusted control of the full-link carbon footprint from raw material procurement to waste recycling through the technical integration of multi-source data collection, dynamic model verification, and blockchain collaborative verification, and solves the core problems of data islands, verification lags, and audit inefficiencies in existing methods.

[0161] Example 1:

[0162] In the production process of chip suppliers in Region A, the edge computing gateway deployed on the lithography machine group collects power consumption data at a sampling frequency of twice per second and transmits it to the blockchain node in real time through the OPC UA protocol. The silicon wafer raw material database purchased by the supplier is accessed through the blockchain oracle interface, and the purchase order is subjected to SHA-256 hashing operation to generate a 128-bit feature value. The zero-knowledge proof verification module verifies the relevance between the feature value and the LCA report provided by the certification agency in Region B. When it is detected that the deviation between the silicon material purity data and the power consumption exceeds 15%, block data with a red trust mark is generated. The intelligent contract engine calls the semiconductor industry benchmark parameter of 0.532 kgCO2 / kWh in Region A to dynamically correct the hourly energy consumption data, and the abnormal data with a correction amplitude exceeding 20% triggers the multi-signature process of the audit node.

[0163] Example 2:

[0164] In the coal-fired boiler of the chemical plant in Region C, thermocouple sensors are deployed to collect steam pressure and coal calorific value data once per minute. Combining with the electronic consignment blockchain provided by the supervision platform, the hazardous waste disposal volume data is obtained through cross-chain relay nodes. When a 4.3% deviation is detected between the monthly coal consumption of 120 tons and the actual combustion volume of 115 tons recorded in the disposal consignment of the boiler, the smart contract automatically locks the carbon emission calculation result of this batch. The dynamic emission factor model generates a range 1 emission coefficient correction value based on the boiler thermal efficiency of 78% and the equipment load rate parameter, and the corrected carbon emission data is written into a specific channel of the Fabric shard chain, with the block generation time interval compressed to 3 seconds.

[0165] Example 3:

[0166] In the transportation task of auto parts from Region D to Region E, the path planning engine generates a theoretically optimal path based on parameters such as a load of 28 tons and a transportation distance of 850 kilometers, and the benchmark carbon emission value is set to 2.1 tons of CO2. The actual driving path deviation mileage collected by the in-vehicle OBD terminal reaches 37 kilometers. After the smart contract detects that the carbon emission increment exceeds the 12% threshold, it triggers a freight settlement freeze instruction. The edge computing node synchronously analyzes the ship AIS data of the port in Region F, extracts parameters such as a cargo volume of 650 TEU and a sailing speed of 14 knots, constructs a carbon emission feature vector for the sea transportation section, and generates an intermodal transportation tracking map after cross-chain binding with the land transportation data. The map update frequency is set to once every 30 minutes.

[0167] Example 4:

[0168] In the battery recycling plant in Region F, temperature and humidity sensors are deployed to collect environmental parameters of the disassembly workshop once every 10 minutes, generating a time-series data sequence bound to the recycling batch BN2024-07. When it is detected that the actual recovered amount of cobalt in ternary lithium batteries reaches 92 kg, exceeding the theoretical threshold of 85 kg marked in the design BOM file and continuously exceeding the limit for 8 hours, the smart contract activates the supplier KPI deduction rule. The energy consumption anomaly detection model analyzes the energy consumption fluctuation characteristics of the air conditioning system during the non-operating period from 1:00 to 5:00 in the morning, generates a credibility score of 72 points and writes it into the verification parameter library, triggering an on-chain alarm instruction to be synchronized to the regulatory node in Region G. After the geographic information system identifies 3 suppliers with more than 15% carbon emission deviation concentrated in Region H, it generates a production capacity compression suggestion and writes it into the smart contract queue.

[0169] The technical features of the present invention are explained as follows:

[0170] Edge computing gateway: Deployed at the IoT terminal of the supplier's production equipment, it is used to collect data such as energy consumption in real time, and has local data processing capabilities to reduce cloud transmission latency. For example, in the chip production process, the power consumption data of the lithography machine group is collected at a sampling frequency of twice per second and transmitted to the blockchain node through an industrial protocol (such as OPC UA).

[0171] Blockchain Network: A distributed ledger system adopting a sharded chain architecture, which stores data separately according to the regions where suppliers are located (for example, the Hyperledger Fabric framework is used in Region A and the FISCO BCOS framework is used in Region B), and synchronizes block header hash values through a cross-chain relay protocol to ensure the global consistency of multi-regional data.

[0172] Data Standardization Module: A processing unit that converts multi-source heterogeneous data (such as power consumption, transportation mileage, recovery rate) into a unified carbon emission equivalent value. It calls an industry benchmark database (such as the emission factor of 0.532 kgCO2 / kWh for the semiconductor industry in Region A) for unit conversion and generates a structured data summary including timestamps and geographical coordinates.

[0173] Dynamic Emission Factor Model: An algorithm model that dynamically adjusts the carbon emission calculation logic based on device operation parameters (such as boiler thermal efficiency of 78% and device load rate). By verifying the logical correlation between process improvement records and energy consumption data, it generates on-chain records with verification identifiers to support the dynamic correction of carbon emission accounting.

[0174] Blockchain Oracle Interface: Middleware that connects to external data sources (such as the raw material procurement database of a third-party certification agency), performs SHA-256 hashing on the original procurement vouchers to generate 128-bit feature values for cross-chain data interaction and verification.

[0175] Zero-Knowledge Proof Verification Module: A functional module that uses cryptographic algorithms to verify the correlation between procurement vouchers and carbon emission inventories. For example, it compares the hash features of silicon wafer procurement with the bill of materials data in the LCA report to verify data consistency without revealing the original data.

[0176] Cross-Chain Relay Node: A communication node that enables data interaction between different blockchain networks (such as the waste disposal chain of the regulatory platform and the enterprise production chain). It verifies e-waybill data through digital signatures, synchronizes hazardous waste disposal volume information to the enterprise chain, and triggers the locking operation of smart contracts for abnormal data.

[0177] Path Planning Algorithm: An algorithm module that generates a theoretically optimal path and a benchmark carbon emission value based on transportation task parameters (load of 28 tons and distance of 850 kilometers). Combining with the actual trajectory data collected by on-vehicle sensors in real time, it calculates the carbon emission increment difference, and triggers an on-chain abnormal event record when the deviation exceeds 12%.

[0178] Metallurgical Recovery Rate Formula: A mathematical model that calculates the theoretical recovery threshold based on the metal element content (such as 85 kg of cobalt element) in the bill of materials of product design, combined with metallurgical process parameters. When the actual extraction amount exceeds the threshold and continues to exceed the limit for 8 hours, it triggers the supplier performance deduction rule.

[0179] Genetic Algorithm Optimization Engine: A calculation module that simulates the biological evolution mechanism to optimize carbon emission factors. Parameters such as the thermal efficiency of production equipment (e.g., 82%) and the energy consumption coefficient of transportation vehicles (e.g., 0.35 L / km) are encoded as chromosome genes, and a candidate factor set is iteratively generated under the constraints of the GHG Protocol range 1 - 3. The optimal parameter combination is selected through voting by blockchain consensus nodes.

[0180] Geographic Information System Spatial Clustering Analysis: A technical means for regional analysis of the carbon emission deviation rate in each link of the supply chain. For example, after identifying that 3 suppliers in region S have a deviation of more than 15%, production capacity adjustment suggestions are generated and written into the intelligent contract execution queue, and the evaluation parameter update is triggered according to the first-in-first-out rule.

[0181] Credibility Score: A verification metric generated based on environmental sensor data (such as warehouse temperature and humidity) and machine learning models. By analyzing the energy consumption fluctuation characteristics during non-operating hours (such as abnormal operation of the air conditioning system from 1:00 to 5:00 in the morning), a score ranging from 0 to 100 is generated and written into the intelligent contract parameter library. A score below 70 triggers the data review process.

[0182] Multi-Party Signature Confirmation Process: A joint verification mechanism for data correction instructions by audit nodes (at least two) in the blockchain network. The elliptic curve digital signature algorithm (such as SM2 / SM9) is used to sign the corrected hash value, and the signature result is anchored to the block header to ensure the legality and traceability of data correction.

[0183] Explanation of GHG Protocol Scope 1 - 3: The GHG Protocol divides carbon emissions into three accounting scopes: Scope 1 covers greenhouse gas emissions generated by production facilities directly controlled by an enterprise (such as CO2 emissions from a factory's coal-fired boiler); Scope 2 involves indirect emissions related to purchased energy (such as carbon emissions from the power plant corresponding to production electricity consumption); Scope 3 includes other indirect emissions generated by activities in the upstream and downstream of the supply chain (such as carbon emissions in raw material transportation, product use, and waste treatment links). In the present invention, constructing a dynamic emission factor model needs to simultaneously meet the composite constraint conditions of Scope 1 (direct production emission verification), Scope 2 (energy consumption carbon footprint accounting), and Scope 3 (logistics transportation and recycling emission tracking). A candidate set of carbon emission factors covering the entire supply chain is generated through the genetic algorithm optimization engine, making the accounting system meet the requirements of international carbon footprint standards.

[0184] The present invention solves the problem of integrating multi-source heterogeneous data by constructing a data acquisition and trustification processing system covering all links of the supply chain. In the production link of suppliers, edge computing gateways are deployed to collect energy consumption in real time. Combined with the blockchain oracle, the raw material procurement data of third-party certification agencies is called. Through hash digest extraction and zero-knowledge proof verification, the relevance between procurement vouchers and carbon emission inventories is ensured. In the manufacturing process data, through the dynamic emission factor model, the logical verification of equipment operation parameters and process improvement records is carried out to generate standardized on-chain records. Logistics transportation data uses the path planning algorithm to generate the theoretical optimal path benchmark value, and combines the real-time working condition data of vehicle-mounted sensors to calculate the incremental difference. When the deviation exceeds the threshold, an abnormal event record is triggered. In the recycling process, the theoretical threshold is calculated based on the metallurgical recovery rate formula, and the actual extraction amount is compared through a smart contract to dynamically adjust the supplier performance parameters.

[0185] The blockchain sharding chain technology realizes the isolated storage and cross-chain synchronous verification of multi-region data. The supplier data in Region A and Region B are written into different blockchain frameworks respectively, and the block header hash values are synchronized through the cross-chain relay protocol to ensure global data consistency. The smart contract engine periodically calls the on-chain verification code, compares the real-time data with the industry benchmark, and triggers off-chain cross-verification by the edge node after an early warning. In the multi-signature confirmation process, the audit node uses the elliptic curve digital signature algorithm to jointly verify the corrected hash value, and the signature result is anchored to the block header to enhance the credibility of data correction.

[0186] The dynamic monitoring mechanism optimizes the weights of the anomaly detection model through the genetic algorithm, combines the spatial clustering analysis of the geographic information system to identify regional centralized carbon emission deviations. When the logistics path changes, the digital twin system generates multiple alternative solutions, and the screening conditions take into account both transportation timeliness and carbon emission reduction goals. The temperature and humidity sensor data in the recycling link cooperate with the machine learning model to identify the characteristics of energy consumption fluctuations during non-operation periods, and generate a credibility score to be written into the smart contract parameter library. The above technical solutions form a "collection-verification-decision" closed loop, realizing the full-link dynamic supervision and precise audit of the supply chain carbon footprint.

Claims

1. A blockchain-based trusted management method for supply chain carbon data, characterized in that Including: Step S101: Obtain the data of all links in the supply chain. The data of all links in the supply chain includes supplier production data, manufacturing process data, logistics transportation data, and recycling and processing data. Among them, the supplier production data collects the real-time energy consumption through the edge computing gateway deployed on the supplier production equipment. The manufacturing process data includes the sensor data of energy equipment in the factory and the process improvement records. The logistics transportation data obtains the real-time operating parameters of the transport vehicle through the vehicle-mounted terminal. The recycling and processing data includes the material recovery rate of product disassembly; Step S102: Input the data of all links in the supply chain into the blockchain network for trust processing. Generate a data digest including spatio-temporal characteristics through the data standardization module, and write the data digest into the blockchain distributed ledger. Among them, the following processing is performed on the supplier production data: Call the regional carbon emission factor in the industry benchmark database to perform data deviation verification, and generate a chain record including a verification identifier; Combine the manufacturing process data with the equipment operating parameters to construct a dynamic emission factor model, and verify the logical relevance between the process improvement record and the energy consumption through the dynamic emission factor model; Step S103: Input the chain record into the intelligent contract execution engine, and make an audit decision based on the preset carbon emission compliance rules. Among them, when it is detected that the verification identifier corresponding to the energy consumption deviates from the industry benchmark threshold, trigger the data correction instruction generation process, and perform multi-party signature confirmation through the blockchain consensus node; Step S104: Generate a supply chain optimization instruction according to the result of the audit decision, and feedback the supply chain optimization instruction to the management end execution device. Among them, the supply chain optimization instruction includes the supplier evaluation parameters and transportation path planning strategies dynamically adjusted based on the anomaly detection result, forming a low-carbon supply chain management closed-loop system that runs through data collection, verification, and decision-making; The step S103 includes: Based on the metal element content parameters marked in the product design bill of materials, calculate the theoretical recovery threshold according to the metallurgical recovery rate formula; Input the actual extraction amount reported by the recycler and the theoretical recovery threshold into the blockchain intelligent contract for deviation comparison; When it is detected that the actual extraction amount exceeds the theoretical threshold and the continuous duration exceeds the preset period, activate the supplier performance deduction rule module in the intelligent contract, and send an alarm instruction including the abnormal batch identifier to the blockchain node with regulatory authority; It also includes: Collect the monitoring data of the warehouse temperature and humidity sensor within the preset time window, and generate an environmental parameter time series sequence bound to the product recovery batch code; Input the environmental parameter time series sequence into the energy consumption anomaly detection model trained based on historical recovery data to identify the energy consumption fluctuation characteristics during non-operation periods; Correlate and match the energy consumption fluctuation characteristics with the time stamp of the recycling and processing data, and generate a credibility score and write it into the verification parameter library of the blockchain intelligent contract.

2. The trusted management method for supply chain carbon data based on blockchain according to claim 1, wherein The step S102 also includes: Call the raw material procurement database of the third-party certification agency through the blockchain oracle interface, extract the hash digest of the original procurement voucher, and generate a hash characteristic value including the time stamp; Input the hash feature value and the bill of materials data in the life cycle assessment report into the zero-knowledge proof verification module to verify the relevance and consistency between the procurement voucher and the carbon emission inventory. Generate a blockchain data block including a trusted identifier based on the verification result, where the trusted identifier is used to update the evaluation status of the integrity of the supplier's production data.

3. The method for trusted management of supply chain carbon data based on blockchain according to claim 1, wherein, The step S102 includes Access the waste disposal blockchain of the regulatory platform through a cross-chain relay node to obtain the digitally signed electronic waybill data. Extract the time-series consumption data of the chemical inventory management system at the production end, and calculate the dynamic deviation rate between the time-series consumption data and the disposal volume of the electronic waybill. When the dynamic deviation rate exceeds the preset threshold, trigger the smart contract to lock the on-chain status of the carbon emission calculation result of the corresponding production batch.

4. The blockchain-based trusted management method for supply chain carbon data according to claim 1, wherein The step S102 includes: Based on the starting or ending coordinates of the transportation task and the vehicle load parameters, call the path planning algorithm to generate the baseline carbon emission value of the theoretically optimal transportation path. Obtain the real-time working condition data of the actual transportation trajectory through in-vehicle sensors, and calculate the incremental difference between the actual carbon emissions and the baseline carbon emission value. When the incremental difference exceeds the path deviation threshold preset in the smart contract, create an abnormal transportation event record in the blockchain ledger and trigger the suspension execution instruction of the on-chain freight settlement contract.

5. The method for trusted management of supply chain carbon data based on blockchain according to claim 4, characterized in that It also includes: Parse the energy consumption log data of the port handling equipment through the edge computing node, and extract the speed and cargo capacity parameters in the ship AIS trajectory data. Synchronously collect the GPS trajectory data and load change parameters of the road transportation vehicle, the freight formation information and energy consumption monitoring data of the railway transportation. Based on the speed and cargo capacity parameters, construct a carbon emission calculation model for the sea transportation section, and integrate the real-time load, path optimization parameters of the road transportation and the energy consumption coefficient of the railway transportation to generate a carbon emission integration model covering multimodal transportation. Extract the carbon emission intensity characteristics of sea, road and railway transportation, and perform associated mapping with the transportation identifier through the preset coding rule. Generate a carbon emission feature vector including ship identifier, road transportation batch code and railway freight waybill number. Perform cross-chain hash binding on the carbon emission feature vector and the blockchain data of the road, railway and sea transportation sections, and generate a carbon footprint tracking map covering multiple transportation modes of road, railway and sea through the smart contract.

6. The method for trusted management of supply chain carbon data based on blockchain according to claim 1, wherein, The step S103 includes: Set at least two consensus nodes with auditing functions in the blockchain network, and configure the carbon emission auditing permission identifier in the node certificate. When the data correction instruction involves the update of the industry benchmark threshold, extract the corrected data characteristics to generate a hash value to be signed. Call the elliptic curve digital signature algorithm through the consensus node to perform joint signature verification on the hash value, and anchor the signature result to the blockchain transaction block header.

7. The blockchain-based trusted management method for supply chain carbon data according to claim 1, characterized in that, It also includes: Based on the real-time carbon emission data of each link in the supply chain, calculate the percentage deviation rate of the carbon emissions of each node from the industry benchmark value. Perform spatial clustering analysis on the percentage deviation rate of the carbon emissions of each node from the industry benchmark value through the geographic information system to identify the characteristics of regional centralized abnormal distribution. Write the production capacity adjustment suggestions including area identification into the intelligent contract execution queue, and the intelligent contract execution queue triggers the update of supplier evaluation parameters according to the first-in, first-out rule.

8. The method for trusted management of supply chain carbon data based on blockchain according to any one of claims 1-7, characterized in that, The construction of the dynamic emission factor model includes: Collect the thermal efficiency value of production equipment, the energy consumption coefficient of transportation vehicles, and the conversion rate of recycled materials through the industrial Internet of Things interface to generate a standardized parameter data set; Input the standardized parameter data set into the genetic algorithm optimization engine, and perform chromosome crossover and mutation iterations under the emission constraint conditions that meet the GHG Protocol Scope 1-3; Conduct compliance voting verification on the generated candidate emission factor set through the blockchain consensus node, and select the parameter combination with a voting passing rate exceeding the preset threshold to update the dynamic emission factor model coefficient.

Images