Supply chain management method based on digital block chain

Abstract

The invention discloses a supply chain management method based on a digital block chain, and the method comprises the following steps: S1, dividing a supply chain participant into four heterogeneous chain networks according to the role type of the supply chain participant; s2, deploying cross-chain relay bridge modules among the heterogeneous chain networks; s3, demand information initiated by manufacturer nodes is collected, and a bilateral matching graph is constructed; s4, performing supply and demand matching by adopting an improved stable matching algorithm to generate a matching result set; s5, deploying a matching contract according to the matching result set, and completing state synchronization and atomic locking through a cross-chain relay bridge module; s6, executing the matching contract, collecting performance data in real time, and mapping the performance data into a credit score; s7, writing the credit score of each matched node into a supervision chain network; and S8, when delay, default or credit decline events occur in the performance data, executing S4 to S6 again, and completing block chain storage. According to the invention, a heterogeneous chain network system is constructed, dynamic matching and a credit evolution mechanism are fused, and intelligent scheduling and supervision of the supply chain are realized.

Description

Technical Field

[0001] The present invention relates to the field of blockchain and supply chain management technology, and in particular to a supply chain management method based on digital blockchain. Background Art

[0002] With the increasing informatization and digitization of the global manufacturing and logistics industries, supply chain management systems have gradually evolved into complex network systems that require multi-party collaboration, dynamic resource scheduling, and timely fulfillment. Traditional supply chain information systems are mostly built on centralized platform architectures, with leading enterprises or intermediary service platforms centrally controlling transaction processes, data flows, and credit assessments. While these centralized models can improve system organizational efficiency to a certain extent, they are increasingly exposed to structural bottlenecks such as data asymmetry, low coordination efficiency, lack of credit mechanisms, and difficulty in tracing accountability, given the large number of participants, severe data heterogeneity, and frequent cross-system collaboration during fulfillment. These bottlenecks are particularly prominent in the integrated manufacturing, logistics, and finance scenarios involving multiple parties.

[0003] In recent years, blockchain technology, with its distributed ledger, immutable data, and full-process traceability, has been widely explored for optimizing supply chain management models. For example, some solutions attempt to build a material flow information sharing platform based on consortium chains, recording core supply chain transactions on-chain to enhance transparency and auditability. Other research incorporates smart contract technology to automatically trigger certain fulfillment conditions and provide non-repudiation records of credit credentials. However, these single-chain architecture-based solutions generally fail to encompass the entire supply chain ecosystem. Their main limitations lie in three key areas: First, the role division is crude, with unclear responsibilities between manufacturers, suppliers, logistics companies, and regulators, and a lack of customizable and autonomous chain deployment mechanisms. Second, the coarse granularity of data exchange prevents high-precision synchronization and verification of states and events between chains, resulting in weak cross-system contract collaboration. Third, the deployment and execution mechanisms of smart contracts are too static to support dynamic adjustments to resource matching strategies based on contract performance, nor can they form an evolving credit mechanism.

[0004] Furthermore, existing supply chain resource matching mechanisms are mostly based on static rules or centralized scheduling models, failing to incorporate the dynamic game-playing characteristics of stable matching theory in economics. This results in a lack of global optimality and fairness in supply and demand matching. Furthermore, credit scoring mechanisms in existing systems are often superficial and lagging, often providing a single score based solely on past transaction results. This fails to implement real-time credit evolution and on-chain version control management for nodes during the fulfillment process. These issues directly impact the stability of multi-party collaborative systems, resource utilization efficiency, and the credibility of credit constraints.

[0005] Furthermore, while existing cross-chain technologies have achieved partial success in digital asset transfers or multi-chain message notifications, their adaptability to supply chain scenarios remains limited. These issues primarily manifest in the following ways: a lack of a state-locking mechanism for the fulfillment process, which fails to guarantee transaction consistency across chains; a failure to design a unified record chain for regulation, which leads to data silos persisting in multi-chain deployments; and, in practical applications, a failure to consider the dynamic adjustment of resource node service saturation and manufacturer priority requirements, resulting in low system scheduling efficiency, poor pairing stability, and frequent redundancy, congestion, and delays in resource matching.

[0006] Therefore, how to provide a supply chain management method based on digital blockchain is an urgent problem that technicians in this field need to solve. Summary of the Invention

[0007] One objective of the present invention is to propose a supply chain management method based on a digital blockchain. This method constructs a heterogeneous blockchain collaborative system consisting of a manufacturer chain network, a supplier chain network, a logistics chain network, and a supervisory chain network. This system integrates an improved stable matching algorithm with an on-chain credit evolution mechanism to accurately achieve dynamic matching of multi-party supply and demand and intelligent supervision of the performance process. Through a cross-chain relay bridge, state synchronization and transaction consistency are achieved, improving matching efficiency and responsiveness, ensuring credible performance, controllable credit, and full-process traceability. This system possesses the technical advantages of being distributed, highly secure, highly flexible, and scalable.

[0008] A supply chain management method based on digital blockchain according to an embodiment of the present invention includes the following steps:

[0009] S1. Based on the role types of supply chain participants, the supply chain network is divided into four heterogeneous chain networks, including manufacturer chain network, supplier chain network, logistics chain network and supervision chain network;

[0010] S2. Deploy a cross-chain relay bridge module between heterogeneous chain networks. The cross-chain relay bridge module performs transaction status confirmation and message synchronization between chain networks based on the Merkle tree structure and hash time lock mechanism;

[0011] S3. Collect demand information initiated by manufacturer nodes, model the manufacturer nodes as demand-side nodes, and model the supplier nodes and logistics nodes as resource-side nodes, and construct a two-sided matching graph;

[0012] S4. Calculate weights based on the historical performance characteristics, transaction amounts, credit scores, and delivery cycles of the edges in the two-sided matching graph, use an improved stable matching algorithm to match supply and demand, and generate a set of matching results.

[0013] S5. Based on the matching result set, deploy matching contracts in the manufacturer chain network, supplier chain network, and logistics chain network respectively, and complete the state synchronization and atomic locking of the matching contracts through the cross-chain relay bridge module;

[0014] S6. Execute the matching contract and collect performance data of the matched manufacturer nodes, supplier nodes, and logistics nodes in real time, and map the performance data into credit scores of the corresponding nodes;

[0015] S7. Write the credit score of each matched node into the custody chain network and update the historical credit score record of the corresponding node in the custody chain network;

[0016] S8. When there is a delay, default or credit decline in the performance data, the rematching process is automatically triggered, and steps S4 to S6 are re-executed and the blockchain storage is completed.

[0017] Optionally, the S2 specifically includes:

[0018] S21. Set up a cross-chain relay bridge module between the manufacturer chain network, the supplier chain network, the logistics chain network, and the supervision chain network. The cross-chain relay bridge module includes a cross-chain verification node, a relay control node, and a status submission interface.

[0019] S22. Construct a snapshot data structure of the chain status in each heterogeneous chain network, and encapsulate the current block header information, the latest transaction summary, and the account status hash of each chain network into a standard Merkle tree structure;

[0020] S23. When a cross-chain transaction is initiated, the relay control node initiates a transaction lock request, submits the hash timing lock condition to the source chain, and generates a transaction lock status;

[0021] S24. Extract the Merkle proof containing the transaction lock status from the source chain using the cross-chain verification node, and broadcast the Merkle proof to the target chain network through the status submission interface;

[0022] S25. After the target chain network receives the Merkle proof, it executes the verification logic and performs a hash consistency check on the transaction status of the source chain. If the verification passes, the state preparation process of the target chain is started;

[0023] S26. After the target chain completes state preparation, a confirmation signal is returned to the relay control node, which releases the hash timing lock and records the successful synchronization mark of the source chain and target chain states of the cross-chain transaction, and stores the synchronization log on the chain in the supervision chain network.

[0024] Optionally, the S3 specifically includes:

[0025] S31. Collect standardized demand information submitted by manufacturer nodes in the manufacturer chain network and generate a structured demand data set;

[0026] S32. Extract resource capability information of each supplier node in the supplier chain network and transportation capability information of each logistics node in the logistics chain network to construct a resource capability vector;

[0027] S33. Based on the structured demand data set and resource capability vector, the manufacturer node, supplier node, and logistics node are functionally divided and marked as demand-side nodes and resource-side nodes respectively;

[0028] S34. Construct a multi-dimensional supply-demand association edge between nodes, wherein the multi-dimensional supply-demand association edge includes attribute fields of historical cooperation frequency, material category matching, credit score similarity, and geographic service radius;

[0029] S35. Generate a two-sided matching graph based on the information of the nodes and edges. The two-sided matching graph is a type of weighted bipartite graph, in which one side is a demand side node set and the other side is a resource side node set.

[0030] Optionally, the standardized demand information specifically includes the required material type, quantity, delivery cycle, expected price and performance requirements.

[0031] Optionally, the S4 specifically includes:

[0032] S41. Calculate the weight value based on the historical performance characteristics, transaction amount, credit score, and delivery cycle of the edge in the two-sided matching graph:

[0033] w ij =α1H ij +α2T ij +α3C ij +α4D ij ;

[0034] Among them, w ij represents the weight value of the edge connecting the i-th manufacturer node and the j-th resource node, H ij represents the historical fulfillment success rate of the i-th manufacturer node and the j-th resource party node, T ij represents the normalized value of the historical average transaction amount between the i-th manufacturer node and the j-th resource node, C ij Indicates the current credit score of the resource node, D ij represents the time matching degree of the delivery cycle, α1, α2, α3 and α4 represent weight coefficients;

[0035] S42. Execute an improved stable matching algorithm on the manufacturer node set and the resource node set;

[0036] S43, iteratively execute the pairing process, for each pair of matching relationships, if d i Make a pairing request and r j Prefer d in the current pairing state i , then replace the current pairing, otherwise maintain the original pairing relationship and generate an intermediate matching pair set M0, where d i represents the i-th manufacturer node, r j represents the j-th resource node;

[0037] S44, perform local adjustment and optimization on the pairing relationships in the intermediate matching pair set M0, exclude nodes that frequently default in the last T cycles, and finally form a matching result set M * ;

[0038] S45. Output matching result set M * And generate a unique contract identification code UID for each pair of pairing relationships in the matching result set.

[0039] Optionally, the S43 specifically includes:

[0040] S431. Calculate the dynamic priority factor of each manufacturer node:

[0041]

[0042] Among them, δ i represents the dynamic priority factor of the i-th manufacturer node, R i represents the remaining demand quantity of the i-th manufacturer node, Q i represents the number of successful matches of the i-th manufacturer node in the last T cycles, ∈ represents the smoothing factor;

[0043] S432. Define a matching load function for each resource node:

[0044]

[0045] Among them, ψ j represents the matching load function of the j-th resource node, exp represents the exponential function, β represents the adjustment factor, and L j represents the cumulative number of matches of the j-th resource node in the current cycle, and θ represents the load threshold;

[0046] S433, the dynamic priority factor δ of each manufacturer node i With weight value w ij Multiply them together to get the preference of each manufacturer node and sort them in ascending order to form a preference order list;

[0047] S434: The manufacturer node sends pairing requests to the resource nodes in the preference order list in sequence. Each manufacturer node initiates at most one request in each round, skipping the rejected or locked nodes.

[0048] S435. Each resource node maintains the current pairing status and, upon receiving a new manufacturer node request, uses the matching load function as a suppression factor and multiplies it by the weight value to obtain a new weight value:

[0049] S436. If the resource node has not been paired, it immediately accepts the manufacturer node's request. If it has been paired, it compares the weight of the existing pairing with the weight of the new requested pairing. If the weight of the new requested pairing is greater than the weight of the existing pairing, the resource node agrees to replace the pairing and cancels the original pairing. Otherwise, it rejects the new pairing request.

[0050] S437. The rejected manufacturer node attempts the next resource node in the preference order list in the next round. Meanwhile, the replaced manufacturer node also enters the next round and re-initiates the request.

[0051] Optionally, the S5 specifically includes:

[0052] S51, parsing matching result set M * , mapping each pair of matching relationships into a set of matching contract parameters, including manufacturer node identifier, resource party node identifier, material requirement items, delivery cycle, credit requirements, and contract amount content;

[0053] S52. Initiate a matching contract deployment request in the manufacturer chain network and construct a master control contract. The master control contract includes a contract identity code, matching parameters, performance requirements, status fields, and event triggers, generating a unique contract instance.

[0054] S53. Deploy mirror contract copies in the supplier chain network and the logistics chain network respectively. The mirror contract copies maintain the same content as the master control contract but limit the role permissions. The supplier node and the logistics node respectively perform the fulfillment response;

[0055] S54. After all mirror contract copies have passed the status proof verification, the cross-chain relay bridge module will trigger the hash timing lock mechanism to freeze the asset status of the corresponding manufacturer node, supplier node, and logistics node in their respective chain networks and enter the atomic lock state;

[0056] S55. Record the atomic locking event. The cross-chain relay bridge module generates a cross-chain synchronization log and writes it into the supervision chain network to complete the cross-chain consistency record of the contract deployment process.

[0057] Optionally, the S6 specifically includes:

[0058] S61. During the execution of the matching contract, the contract interface is called from the manufacturer chain network, supplier chain network, and logistics chain network respectively to obtain the transaction records of the matched nodes in the current performance cycle;

[0059] S62. Structural processing is performed on the collected transaction execution records to extract the performance data of each node, including the performance start time, performance completion time, material delivery quantity, delivery location, and delivery integrity indicator;

[0060] S63. Parse the event logs during the performance process to identify any abnormal states such as overdue delivery, early performance, interrupted performance, and partial performance, and generate an event tag set.

[0061] S64. Calculate the node's credit score by taking the performance data and the event tag set as input and combining them with the preset credit assessment criteria;

[0062] S65. Classify and mark the credit scores according to different node roles, and classify them into manufacturer node performance files, supplier node performance files, and logistics node performance files respectively;

[0063] S66. Bind the archived credit score to the node's unique identity, generate a credit score record for the current period, and encapsulate it into a structured data block.

[0064] Optionally, the S7 specifically includes:

[0065] S71. Encrypt and encapsulate the credit score record to construct a credit score data packet containing the node identity, score timestamp, score period identifier, score content summary, and previous block reference hash;

[0066] S72. Calling the credit write interface in the custody chain network to write the credit score data packet to the credit status record area of ​​the corresponding node in the custody chain, and generating a unique write transaction number;

[0067] S73. Check whether the node's historical credit score record already exists in the chain of custody network. If so, add the new score entry and mark it as the current valid version, and mark the previous version status as a historical archive status.

[0068] S74. Synchronously update the latest credit index value and version index pointer of the online node of the custody chain.

[0069] The beneficial effects of the present invention are:

[0070] First, the present invention divides the supply chain network into four heterogeneous chain networks: manufacturer chain network, supplier chain network, logistics chain network, and supervision chain network. In combination with their respective business roles, independent chain environments and consensus mechanisms are deployed for each network, thereby achieving data autonomy, authority separation, and clear role responsibilities in the multi-party collaboration of the supply chain. This provides a structural foundation for solving the problems of mixed information flows and extensive authority control among traditional multi-party networks. By deploying a cross-chain relay bridge module based on the Merkle tree structure and hash time lock mechanism, efficient and reliable data status synchronization and transaction coordination between chains are achieved, ensuring data consistency and atomicity during cross-chain interaction, and greatly improving the operational stability and transaction transparency of the cross-chain supply chain system.

[0071] Secondly, the present invention proposes a two-sided matching model based on a weighted bipartite graph constructed based on standardized demand information and resource capability vectors. This model calculates the weights of the edges connecting nodes by integrating multi-dimensional factors such as historical performance characteristics, transaction amounts, credit scores, and delivery cycles. It also introduces dynamic priority adjustment and load control mechanisms to improve the stable matching algorithm, significantly enhancing the fairness of supply and demand matching and system resource utilization. Through pairing request iteration, load suppression, and local adjustment optimization, the present invention achieves a highly efficient and stable dynamic matching process, addressing resource tilt, node overload, and irrational scheduling in traditional supply chains, and enhancing scheduling adaptability in complex multi-node environments.

[0072] In addition, during the contract execution process, the present invention collects performance behavior data from manufacturers, suppliers, and logistics nodes, and combines it with an abnormal event tag identification mechanism for structured processing to build a behavior-based on-chain credit scoring system. This system not only supports real-time updates and version control of credit scores, but also realizes node archiving of performance files and on-chain writing of the supervision chain network, effectively making up for the shortcomings of traditional scoring mechanisms such as poor timeliness, low transparency, and non-traceability. When abnormal performance occurs, the system can automatically identify delays, defaults, or credit decline events, trigger a rematching process, automatically replace bad nodes, and rebuild contractual relationships, ensuring the continuity, flexibility, and security of the overall coordination of the supply chain.

[0073] Ultimately, by uniformly encapsulating and archiving all fulfillment process data, matching relationships, credit scores, and rematching records through the chain of custody network, this invention achieves a complete, traceable, and verifiable on-chain regulatory path, enhances data compliance and responsibility division capabilities, and promotes the transformation of the supply chain from traditional centralized dependent coordination to a distributed autonomous collaborative model. Overall, this invention not only improves the response speed and resource adaptation accuracy of the entire supply chain process, but also enhances the system's adaptability and self-regulation capabilities in complex dynamic environments. It has strong practicality, scalability, and wide-ranging value. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0075] Figure 1 This is an overall flow chart of a supply chain management method based on digital blockchain proposed by the present invention;

[0076] Figure 2 This is a schematic diagram of the heterogeneous chain network architecture of a supply chain management method based on digital blockchain proposed in the present invention;

[0077] Figure 3 This is a flowchart of the execution of an improved stable matching algorithm for a supply chain management method based on digital blockchain proposed in the present invention. DETAILED DESCRIPTION

[0078] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0079] refer to Figure 1-3 , a supply chain management method based on digital blockchain, comprising the following steps:

[0080] S1. Based on the role types of supply chain participants, the supply chain network is divided into four heterogeneous chain networks, including manufacturer chain network, supplier chain network, logistics chain network and supervision chain network;

[0081] S2. Deploy a cross-chain relay bridge module between heterogeneous chain networks. The cross-chain relay bridge module performs transaction status confirmation and message synchronization between chain networks based on the Merkle tree structure and hash time lock mechanism;

[0082] S3. Collect demand information initiated by manufacturer nodes, model the manufacturer nodes as demand-side nodes, and model the supplier nodes and logistics nodes as resource-side nodes, and construct a two-sided matching graph;

[0083] S4. Calculate weights based on the historical performance characteristics, transaction amounts, credit scores, and delivery cycles of the edges in the two-sided matching graph, use an improved stable matching algorithm to match supply and demand, and generate a set of matching results.

[0084] S5. Based on the matching result set, deploy matching contracts in the manufacturer chain network, supplier chain network, and logistics chain network respectively, and complete the state synchronization and atomic locking of the matching contracts through the cross-chain relay bridge module;

[0085] S6. Execute the matching contract and collect performance data of the matched manufacturer nodes, supplier nodes, and logistics nodes in real time, and map the performance data into credit scores of the corresponding nodes;

[0086] S7. Write the credit score of each matched node into the custody chain network and update the historical credit score record of the corresponding node in the custody chain network;

[0087] S8. When there is a delay, default or credit decline in the performance data, the rematching process is automatically triggered, and steps S4 to S6 are re-executed and the blockchain storage is completed.

[0088] The present invention divides the supply chain network into four heterogeneous blockchain systems: manufacturer chain network, supplier chain network, logistics chain network, and supervision chain network, and combines them with a cross-chain relay bridge module to build a multi-role distributed autonomous collaborative management system. This method not only achieves the atomic state synchronization of transactions between chains, but also establishes a resource matching mechanism that combines a bilateral matching graph with an improved stable matching algorithm, thereby improving the rationality and stability of supply and demand pairing. On this basis, performance data can be collected in real time and a dynamic credit score can be generated. After being stored on the chain, it can be continuously tracked and fed back by the supervision chain. In particular, when delays or breaches of contract occur during the performance process, the system can automatically trigger the re-matching process to ensure the timely repair of the supply and demand relationship and the self-healing ability of the system, comprehensively improving the stability, responsiveness and controllability of the supply chain in complex collaborative scenarios.

[0089] In this embodiment, S2 specifically includes:

[0090] S21. Set up a cross-chain relay bridge module between the manufacturer chain network, the supplier chain network, the logistics chain network, and the supervision chain network. The cross-chain relay bridge module includes a cross-chain verification node, a relay control node, and a status submission interface.

[0091] S22. Construct a snapshot data structure of the chain status in each heterogeneous chain network, and encapsulate the current block header information, the latest transaction summary, and the account status hash of each chain network into a standard Merkle tree structure;

[0092] S23. When a cross-chain transaction is initiated, the relay control node initiates a transaction lock request, submits the hash timing lock condition to the source chain, and generates a transaction lock status;

[0093] S24. Extract the Merkle proof containing the transaction lock status from the source chain using the cross-chain verification node, and broadcast the Merkle proof to the target chain network through the status submission interface;

[0094] S25. After the target chain network receives the Merkle proof, it executes the verification logic and performs a hash consistency check on the transaction status of the source chain. If the verification passes, the state preparation process of the target chain is started;

[0095] S26. After the target chain completes state preparation, a confirmation signal is returned to the relay control node, which releases the hash timing lock and records the successful synchronization mark of the source chain and target chain states of the cross-chain transaction, and stores the synchronization log on the chain in the supervision chain network.

[0096] By designing the structure and operation mechanism of the cross-chain relay bridge module in detail, we have achieved technical guarantees for data interaction and state consistency between heterogeneous chain networks. Through the combination of cross-chain verification nodes, relay control nodes, and state submission interfaces, in conjunction with the Merkle tree structure and hash timing lock mechanism, this module can complete the generation of state snapshots between chains, atomic locking of cross-chain transactions, multi-party broadcast verification of state submissions, and final consistency, effectively solving the core problems of large cross-chain interaction delays, inconsistent states, and irreversible transactions in existing multi-chain systems. Its design enables trusted cross-chain collaboration between any two business chain networks in the supply chain, improves the efficiency of on-chain data synchronization and the overall stability of the system, and is highly portable and modular, making it easy to expand and implement.

[0097] In this embodiment, S3 specifically includes:

[0098] S31. Collect standardized demand information submitted by manufacturer nodes in the manufacturer chain network and generate a structured demand data set;

[0099] S32. Extract resource capability information of each supplier node in the supplier chain network and transportation capability information of each logistics node in the logistics chain network to construct a resource capability vector;

[0100] S33. Based on the structured demand data set and resource capability vector, the manufacturer node, supplier node, and logistics node are functionally divided and marked as demand-side nodes and resource-side nodes respectively;

[0101] S34. Construct a multi-dimensional supply-demand association edge between nodes, wherein the multi-dimensional supply-demand association edge includes attribute fields of historical cooperation frequency, material category matching, credit score similarity, and geographic service radius;

[0102] S35. Generate a two-sided matching graph based on the information of the nodes and edges. The two-sided matching graph is a type of weighted bipartite graph, in which one side is a demand side node set and the other side is a resource side node set.

[0103] This step uses a standardized and structured approach to uniformly model the supply and demand data of manufacturers, suppliers, and logistics nodes, and on this basis constructs resource capability vectors and demand vectors, thus realizing a graph model representation of the demand side and the resource side. The established multi-dimensional supply and demand association edges integrate multiple key indicators such as historical cooperation frequency, category matching, credit score, and service radius, improving the semantic expression ability and business adaptability of the matching graph structure. The two-sided matching graph is constructed by means of a weighted bipartite graph, providing a structural optimization and high-quality input foundation for the subsequent stable matching algorithm. This modeling method is significantly superior to traditional matching methods based on single field screening or rule table drive, with stronger scalability, upstream and downstream semantic fidelity, and dynamic adaptability, providing a new technical paradigm for multi-party supply chain matching relationship modeling.

[0104] In this embodiment, the standardized demand information specifically includes the required material type, quantity, delivery cycle, expected price and performance requirements.

[0105] In this embodiment, the S4 specifically includes:

[0106] S41. Calculate the weight value based on the historical performance characteristics, transaction amount, credit score, and delivery cycle of the edge in the two-sided matching graph:

[0107] w ij =α1H ij +α2T ij +α3C ij +α4D ij ;

[0108] Among them, w ij represents the weight value of the edge connecting the i-th manufacturer node and the j-th resource node, H ij represents the historical fulfillment success rate of the i-th manufacturer node and the j-th resource party node, T ij represents the normalized value of the historical average transaction amount between the i-th manufacturer node and the j-th resource node, C ij Indicates the current credit score of the resource node, D ij represents the time matching degree of the delivery cycle, α1, α2, α3 and α4 represent weight coefficients;

[0109] S42. Execute an improved stable matching algorithm on the manufacturer node set and the resource node set;

[0110] S43, iteratively execute the pairing process, for each pair of matching relationships, if d i Make a pairing request and r j Prefer d in the current pairing state i , then replace the current pairing, otherwise maintain the original pairing relationship and generate an intermediate matching pair set M0, where di represents the i-th manufacturer node, r j represents the j-th resource node;

[0111] S44, perform local adjustment and optimization on the pairing relationships in the intermediate matching pair set M0, exclude nodes that frequently default in the last T cycles, and finally form a matching result set M * ;

[0112] S45. Output matching result set M * And generate a unique contract identification code UID for each pair of pairing relationships in the matching result set.

[0113] By weighting the edges in the bilateral matching graph based on multiple factors such as fulfillment history, transaction amount, credit score, and delivery cycle, the comprehensive representativeness of the edge weights is improved, and this is used as the core input of the matching algorithm to provide a multi-dimensional quantitative basis for subsequent matching decisions. In the matching process, dynamic optimization is performed by combining historical behavior and business expectations, which can effectively screen the optimal resource path and minimize resource tilt and scheduling conflicts. At the same time, the unique contract identification code UID is introduced to achieve precise binding of each pair of matching relationships with the on-chain contract deployment, ensuring that the fulfillment process is traceable and auditable. Overall, this step significantly improves the accuracy, fairness, and consistency of on-chain execution of matching, and promotes the evolution of the traditional resource selection process to an intelligent game type.

[0114] In this embodiment, the S43 specifically includes:

[0115] S431. Calculate the dynamic priority factor of each manufacturer node:

[0116]

[0117] Among them, δ i represents the dynamic priority factor of the i-th manufacturer node, R i represents the remaining demand quantity of the i-th manufacturer node, Q i represents the number of successful matches of the i-th manufacturer node in the last T cycles, ∈ represents the smoothing factor;

[0118] S432. Define a matching load function for each resource node:

[0119]

[0120] Among them, ψ j represents the matching load function of the j-th resource node, exp represents the exponential function, β represents the adjustment factor, and L j represents the cumulative number of matches of the j-th resource node in the current cycle, and θ represents the load threshold;

[0121] S433, the dynamic priority factor δ of each manufacturer node i With weight value w ij Multiply them together to get the preference of each manufacturer node and sort them in ascending order to form a preference order list;

[0122] S434: The manufacturer node sends pairing requests to the resource nodes in the preference order list in sequence. Each manufacturer node initiates at most one request in each round, skipping the rejected or locked nodes.

[0123] S435. Each resource node maintains the current pairing status and, upon receiving a new manufacturer node request, uses the matching load function as a suppression factor and multiplies it by the weight value to obtain a new weight value:

[0124] S436. If the resource node has not been paired, it immediately accepts the manufacturer node's request. If it has been paired, it compares the weight of the existing pairing with the weight of the new requested pairing. If the weight of the new requested pairing is greater than the weight of the existing pairing, the resource node agrees to replace the pairing and cancels the original pairing. Otherwise, it rejects the new pairing request.

[0125] S437. The rejected manufacturer node attempts the next resource node in the preference order list in the next round. Meanwhile, the replaced manufacturer node also enters the next round and re-initiates the request.

[0126] The above steps elaborate on the calculation process and key parameter control mechanism of the improved stable matching algorithm, including the introduction of dynamic priority factors and load suppression functions, which solves the problem of poor stability of the traditional Gale-Shapley algorithm under conditions of resource constraints and uneven requests. The manufacturer node dynamically adjusts the matching priority based on the remaining demand and historical pairing situation, and the resource node dynamically suppresses the request response based on the matching frequency of the current cycle, effectively avoiding the concentration of resources on individual nodes. During the entire matching process, the pairing relationship is gradually optimized in multiple rounds of iterations, and finally a stable, efficient, and balanced supply and demand scheduling state is achieved. This improved algorithm is suitable for supply chain scenarios with multi-role and multi-state constraints. It has obvious advantages in adaptability and scheduling flexibility, and is significantly better than existing static strategies.

[0127] In this embodiment, the S5 specifically includes:

[0128] S51, parsing matching result set M * , mapping each pair of matching relationships into a set of matching contract parameters, including manufacturer node identifier, resource party node identifier, material requirement items, delivery cycle, credit requirements, and contract amount content;

[0129] S52. Initiate a matching contract deployment request in the manufacturer chain network and construct a master control contract. The master control contract includes a contract identity code, matching parameters, performance requirements, status fields, and event triggers, generating a unique contract instance.

[0130] S53. Deploy mirror contract copies in the supplier chain network and the logistics chain network respectively. The mirror contract copies maintain the same content as the master control contract but limit the role permissions. The supplier node and the logistics node respectively perform the fulfillment response;

[0131] S54. After all mirror contract copies have passed the status proof verification, the cross-chain relay bridge module will trigger the hash timing lock mechanism to freeze the asset status of the corresponding manufacturer node, supplier node, and logistics node in their respective chain networks and enter the atomic lock state;

[0132] S55. Record the atomic locking event. The cross-chain relay bridge module generates a cross-chain synchronization log and writes it into the supervision chain network to complete the cross-chain consistency record of the contract deployment process.

[0133] The present invention proposes a multi-chain contract collaborative execution structure consisting of a master control contract and a mirror contract, and realizes atomic deployment control through cross-chain state locking and hash timing mechanisms. The matching contract deployment parameters are refined into node identification, performance parameters and constraints to ensure that the semantics of the smart contract are strictly consistent with the actual transaction business. The execution authority is limited by the mirror contract, corresponding to the manufacturer, supplier and logistics roles respectively, effectively controlling the execution boundary and operation authorization to avoid contract abuse or misoperation. Finally, the contract deployment behavior and status synchronization log are uniformly recorded through the supervision chain network to ensure that the entire execution process is transparent and traceable, providing a stable and efficient deployment mechanism for contract collaboration in heterogeneous chain environments.

[0134] In this embodiment, S6 specifically includes:

[0135] S61. During the execution of the matching contract, the contract interface is called from the manufacturer chain network, supplier chain network, and logistics chain network respectively to obtain the transaction records of the matched nodes in the current performance cycle;

[0136] S62. Structural processing is performed on the collected transaction execution records to extract the performance data of each node, including the performance start time, performance completion time, material delivery quantity, delivery location, and delivery integrity indicator;

[0137] S63. Parse the event logs during the performance process to identify any abnormal states such as overdue delivery, early performance, interrupted performance, and partial performance, and generate an event tag set.

[0138] S64. Calculate the node's credit score by taking the performance data and the event tag set as input and combining them with the preset credit assessment criteria;

[0139] S65. Classify and mark the credit scores according to different node roles, and classify them into manufacturer node performance files, supplier node performance files, and logistics node performance files respectively;

[0140] S66. Bind the archived credit score to the node's unique identity, generate a credit score record for the current period, and encapsulate it into a structured data block.

[0141] This step refines the on-chain collection of performance data, anomaly identification, and credit score calculation process, and realizes the automatic conversion of node performance into structured credit records. The system can extract performance behavior logs from multiple chains in real time, and generate multi-dimensional performance labels based on timestamps, delivery status, and abnormal events, thereby forming a quantitative credit score based on behavioral evidence. This scoring mechanism has the advantages of strong reusability, fast response speed, and high upstream and downstream consistency. It also provides computable dynamic input indicators for subsequent rematching strategies. Compared with existing static credit models, the present invention has significant improvements in scoring timeliness, scoring structure rationality, and scoring scenario adaptability.

[0142] In this embodiment, the S7 specifically includes:

[0143] S71. Encrypt and encapsulate the credit score record to construct a credit score data packet containing the node identity, score timestamp, score period identifier, score content summary, and previous block reference hash;

[0144] S72. Calling the credit write interface in the custody chain network to write the credit score data packet to the credit status record area of ​​the corresponding node in the custody chain, and generating a unique write transaction number;

[0145] S73. Check whether the node's historical credit score record already exists in the chain of custody network. If so, add the new score entry and mark it as the current valid version, and mark the previous version status as a historical archive status.

[0146] S74. Synchronously update the latest credit index value and version index pointer of the online node of the custody chain.

[0147] This step establishes an on-chain mechanism for writing, versioning, and indexing credit score records. By encrypting and encapsulating the credit score for each cycle, a scoring data package containing node identification, timestamp, and summary information is generated. This, combined with the Chain of Custody network write interface, enables on-chain visualization of node-level credit status. Historical versions are archived via index pointers, while current versions are dynamically updated throughout the fulfillment cycle, creating a scalable, queryable, and traceable repository of credit evolution trajectories. This design significantly enhances the structural integrity, traceability, and risk control capabilities of credit scores, meeting the multi-dimensional requirements for fulfillment and risk management in complex supply chain scenarios.

[0148] Example 1:

[0149] To verify the feasibility of this invention, it was applied to a high-end electronics manufacturing supply chain collaboration platform serving over 80 electronic component suppliers, 15 large assembly manufacturers, and over 10 logistics providers in the region. Led by a Shenzhen-based technology company, the project aims to address core issues that have long plagued collaboration among regional enterprises, including low multi-node order collaboration efficiency, irrational resource allocation, and the lack of transparent sharing of fulfillment credit data.

[0150] In this embodiment, the original system relies on a centralized ERP platform for order matching and supplier screening. It lacks a real-time credit feedback mechanism and on-chain traceability. Matching results are often uniformly matched by the platform, resulting in low resource utilization and unstable matching response time, which cannot meet the increasingly complex needs of customized small-batch rapid delivery. After introducing this invention, the platform first deploys four types of heterogeneous chain networks based on the type of participants, including manufacturer chain network, supplier chain network, logistics chain network, and supervision chain network. By deploying a cross-chain relay bridge module based on Merkle tree and hash time lock, it achieves high-reliability synchronization of inter-chain transactions.

[0151] During the system's trial run, a total of 62 enterprise nodes participated in on-chain order scheduling, including 12 manufacturers, 41 suppliers, and 9 logistics providers. The average daily order request volume was 286. The system automatically mapped the order requirements submitted by all manufacturers and the resource nodes using the two-sided weighted matching graph model of this invention. During the matching process, a dynamic priority mechanism was used to control the order sequence of manufacturer requests, and the resource node's willingness to accept was dynamically adjusted based on its current cycle service saturation. This significantly shortened the matching response time. Taking the second week of January 2025 as an example, the average matching completion time was 8.1 seconds, a 63.9% reduction from the average 22.5 seconds of the old system.

[0152] During the actual fulfillment process, the system records the initiation time, matching node, delivery time, deviation from standards, delivery completeness, and dispute events of each order, and generates a credit score based on this information. The score of each node is updated in the chain of custody network, and the rematching process is triggered according to the score change. During this period, a total of 34 active rematches were triggered, of which 17 were triggered due to credit declines at logistics nodes, 11 were triggered due to overdue delivery at supply nodes, and 6 were triggered due to changes in manufacturer requirements. After rematching, the average delayed fulfillment rate dropped from 12.7% to 4.3%, significantly improving the collaborative resilience of the entire supply network.

[0153] In terms of supervision, the system supports structured packaging and audit upload of all on-chain matching, performance, scoring, and abnormal events. Each supervisory node can quickly search and verify the performance of a specified enterprise node within a specified period, with an average response time of less than 3 seconds.

[0154] Overall, the present invention performs well in high-frequency orders, complex resource scheduling, credit evolution control, and cross-chain regulatory coordination. It combines real-time, stability, and auditability, and is significantly superior to the traditional centralized supply chain scheduling model, making it scalable.

[0155] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A supply chain management method based on digital blockchain, characterized in that: The steps include: S1. Based on the role types of supply chain participants, the supply chain network is divided into four heterogeneous chain networks, including manufacturer chain network, supplier chain network, logistics chain network and supervision chain network; S2. Deploy a cross-chain relay bridge module between heterogeneous chain networks. The cross-chain relay bridge module performs transaction status confirmation and message synchronization between chain networks based on the Merkle tree structure and hash time lock mechanism; S3. Collect demand information initiated by manufacturer nodes, model the manufacturer nodes as demand-side nodes, and model the supplier nodes and logistics nodes as resource-side nodes, and construct a two-sided matching graph; S4. Calculate weights based on the historical performance characteristics, transaction amounts, credit scores, and delivery cycles of the edges in the two-sided matching graph, use an improved stable matching algorithm to match supply and demand, and generate a set of matching results. S5. Based on the matching result set, deploy matching contracts in the manufacturer chain network, supplier chain network, and logistics chain network respectively, and complete the state synchronization and atomic locking of the matching contracts through the cross-chain relay bridge module; S6. Execute the matching contract and collect performance data of the matched manufacturer nodes, supplier nodes, and logistics nodes in real time, and map the performance data into credit scores of the corresponding nodes; S7. Write the credit score of each matched node into the custody chain network, and update the historical credit score record of the corresponding node in the custody chain network; S8. When there is a delay, default or credit decline in the performance data, the rematching process is automatically triggered, steps S4 to S6 are re-executed, and blockchain storage is completed.

2. A supply chain management method based on digital blockchain according to claim 1, characterized in that: The S2 specifically includes: S21. Set up a cross-chain relay bridge module between the manufacturer chain network, the supplier chain network, the logistics chain network, and the supervision chain network. The cross-chain relay bridge module includes a cross-chain verification node, a relay control node, and a status submission interface. S22. Construct a snapshot data structure of the chain status in each heterogeneous chain network, and encapsulate the current block header information, the latest transaction summary, and the account status hash of each chain network into a standard Merkle tree structure; S23. When a cross-chain transaction is initiated, the relay control node initiates a transaction lock request, submits the hash timing lock condition to the source chain, and generates a transaction lock status; S24. Extract the Merkle proof containing the transaction lock status from the source chain using the cross-chain verification node, and broadcast the Merkle proof to the target chain network through the status submission interface; S25. After the target chain network receives the Merkle proof, it executes the verification logic and performs a hash consistency check on the transaction status of the source chain. If the verification passes, the state preparation process of the target chain is started; S26. After the target chain completes state preparation, a confirmation signal is returned to the relay control node, which releases the hash timing lock and records the successful synchronization mark of the source chain and target chain states of the cross-chain transaction, and stores the synchronization log on the chain in the supervision chain network.

3. A supply chain management method based on digital blockchain according to claim 1, characterized in that: The S3 specifically includes: S31. Collect standardized demand information submitted by manufacturer nodes in the manufacturer chain network and generate a structured demand data set; S32. Extract resource capability information of each supplier node in the supplier chain network and transportation capability information of each logistics node in the logistics chain network to construct a resource capability vector; S33. Based on the structured demand data set and resource capability vector, the manufacturer node, supplier node, and logistics node are functionally divided and marked as demand-side nodes and resource-side nodes respectively; S34. Construct a multi-dimensional supply-demand association edge between nodes, wherein the multi-dimensional supply-demand association edge includes attribute fields of historical cooperation frequency, material category matching, credit score similarity, and geographic service radius; S35. Generate a two-sided matching graph based on the information of the nodes and edges. The two-sided matching graph is a type of weighted bipartite graph, in which one side is a demand side node set and the other side is a resource side node set.

4. A supply chain management method based on digital blockchain according to claim 3, characterized in that: The standardized demand information specifically includes the required material type, quantity, delivery cycle, expected price and performance requirements.

5. A supply chain management method based on digital blockchain according to claim 1, characterized in that: The S4 specifically includes: S41. Calculate the weight value based on the historical performance characteristics, transaction amount, credit score, and delivery cycle of the edge in the two-sided matching graph: w ij =α1H ij +α2T ij +α3C ij +α4D ij ; Among them, w ij represents the weight value of the edge connecting the i-th manufacturer node and the j-th resource node, H ij represents the historical fulfillment success rate of the i-th manufacturer node and the j-th resource party node, T ij represents the normalized value of the historical average transaction amount between the i-th manufacturer node and the j-th resource node, C ij Indicates the current credit score of the resource node, D ij represents the time matching degree of the delivery cycle, α1, α2, α3 and α4 represent weight coefficients; S42. Execute an improved stable matching algorithm on the manufacturer node set and the resource node set; S43, iteratively execute the pairing process, for each pair of matching relationships, if d i Make a pairing request and r j Prefer d in the current pairing state i , then replace the current pairing, otherwise maintain the original pairing relationship and generate an intermediate matching pair set M0, where d i represents the i-th manufacturer node, r j represents the j-th resource node; S44, perform local adjustment and optimization on the pairing relationships in the intermediate matching pair set M0, exclude nodes that frequently default in the last T cycles, and finally form a matching result set M * ; S45. Output matching result set M * And generate a unique contract identification code UID for each pair of pairing relationships in the matching result set.

6. A supply chain management method based on digital blockchain according to claim 5, characterized in that: The S43 specifically includes: S431. Calculate the dynamic priority factor of each manufacturer node: Among them, δ i represents the dynamic priority factor of the i-th manufacturer node, R i represents the remaining demand quantity of the i-th manufacturer node, Q i represents the number of successful matches of the i-th manufacturer node in the last T cycles, ∈ represents the smoothing factor; S432. Define a matching load function for each resource node: Among them, ψ j represents the matching load function of the j-th resource node, exp represents the exponential function, β represents the adjustment factor, and L j represents the cumulative number of matches of the j-th resource node in the current cycle, and θ represents the load threshold; S433, the dynamic priority factor δ of each manufacturer node i With weight value w ij Multiply them together to get the preference of each manufacturer node and sort them in ascending order to form a preference order list; S434: The manufacturer node sends pairing requests to the resource nodes in the preference order list in sequence. Each manufacturer node initiates at most one request in each round, skipping the rejected or locked nodes. S435. Each resource node maintains the current pairing status and, upon receiving a new manufacturer node request, uses the matching load function as a suppression factor and multiplies it by the weight value to obtain a new weight value: S436. If the resource node has not been paired, it immediately accepts the manufacturer node's request. If it has been paired, it compares the weight of the existing pairing with the weight of the new requested pairing. If the weight of the new requested pairing is greater than the weight of the existing pairing, the resource node agrees to replace the pairing and cancels the original pairing. Otherwise, it rejects the new pairing request. S437. The rejected manufacturer node attempts the next resource node in the preference order list in the next round. Meanwhile, the replaced manufacturer node also enters the next round and re-initiates the request.

7. A supply chain management method based on digital blockchain according to claim 1, characterized in that: The S5 specifically includes: S51, parsing matching result set M * , mapping each pair of matching relationships into a set of matching contract parameters, including manufacturer node identifier, resource party node identifier, material requirement items, delivery cycle, credit requirements, and contract amount content; S52. Initiate a matching contract deployment request in the manufacturer chain network and construct a master control contract. The master control contract includes a contract identity code, matching parameters, performance requirements, status fields, and event triggers, generating a unique contract instance. S53. Deploy mirror contract copies in the supplier chain network and the logistics chain network respectively. The mirror contract copies maintain the same content as the master control contract but limit the role permissions. The supplier node and the logistics node respectively perform the fulfillment response; S54. After all mirror contract copies have passed the status proof verification, the cross-chain relay bridge module will trigger the hash timing lock mechanism to freeze the asset status of the corresponding manufacturer node, supplier node, and logistics node in their respective chain networks and enter the atomic lock state; S55. Record the atomic locking event. The cross-chain relay bridge module generates a cross-chain synchronization log and writes it into the supervision chain network to complete the cross-chain consistency record of the contract deployment process.

8. A supply chain management method based on digital blockchain according to claim 1, characterized in that: The S6 specifically includes: S61. During the execution of the matching contract, the contract interface is called from the manufacturer chain network, supplier chain network, and logistics chain network respectively to obtain the transaction records of the matched nodes in the current performance cycle; S62. Structural processing is performed on the collected transaction execution records to extract the performance data of each node, including the performance start time, performance completion time, material delivery quantity, delivery location, and delivery integrity indicator; S63. Parse the event logs during the performance process to identify any abnormal states such as overdue delivery, early performance, interrupted performance, and partial performance, and generate an event tag set. S64. Calculate the node's credit score by taking the performance data and the event tag set as input and combining them with the preset credit assessment criteria; S65. Classify and mark the credit scores according to different node roles, and classify them into manufacturer node performance files, supplier node performance files, and logistics node performance files respectively; S66. Bind the archived credit score to the node's unique identity, generate a credit score record for the current period, and encapsulate it into a structured data block.

9. A supply chain management method based on digital blockchain according to claim 1, characterized in that: The S7 specifically includes: S71. Encrypt and encapsulate the credit score record to construct a credit score data packet containing the node identity, score timestamp, score period identifier, score content summary, and previous block reference hash; S72. Calling the credit write interface in the custody chain network to write the credit score data packet to the credit status record area of ​​the corresponding node in the custody chain, and generating a unique write transaction number; S73. Check whether the node's historical credit score record already exists in the chain of custody network. If so, add the new score entry and mark it as the current valid version, and mark the previous version status as a historical archive status. S74. Synchronously update the latest credit index value and version index pointer of the online node of the custody chain.

Images