A Blockchain-Based Approach to Efficient Collaborative Management of Supply Chain Nodes

Abstract

This invention discloses a blockchain-based method for efficient collaborative management and control of supply chain nodes, belonging to the field of supply chain management. It achieves a unified centralized real-time linkage and forward-looking collaboration by constructing a blockchain network, establishing inter-node dependencies, implementing two-way chain effect assessment, and forward-looking plan correction. On the one hand, it bidirectionally searches related nodes along the resource dependence direction of anomaly sources, independently quantifies the disturbance diffusion intensity of each anomaly source, and superimposes multiple sources to locate the disturbance impact range threshold. It generates targeted pre-adjustment instructions for nodes within the domain, overcoming the problem of rigid collaborative mechanisms and difficulty in quickly responding to multi-point linkage changes caused by neglecting the interaction constraints between nodes in traditional solutions. On the other hand, based on future deviation prediction, it enables each node to perceive the chain effect of upstream and downstream estimated deviations and make advance plan adjustments, overcoming the passive response limitations of traditional technologies that can only provide post-event warnings and cannot analyze future risks in advance.

Description

Technical Field

[0001] This invention relates to the field of supply chain management, and more specifically, to a method for efficient collaborative management and control of supply chain nodes based on blockchain analysis. Background Technology

[0002] A supply chain network is a complex and dynamic system driven by the interaction of material flow, information flow, and capital flow. Supply chain nodes refer to the various independent participants or key business links that constitute the supply chain network. As the basic units constituting the supply chain network (such as suppliers, manufacturers, logistics providers, distributors, etc.), their collaborative efficiency directly determines the response speed and operating costs of the entire supply chain.

[0003] Traditional supply chain management solutions, such as centralized planning systems based on Enterprise Resource Planning (ERP) or Electronic Data Interchange (EDI), have long been plagued by problems such as information silos, lack of trust, and the bullwhip effect. To address supply chain security issues, existing technologies have also proposed solutions for risk management of supply chain nodes. For example, Chinese patent application CN117499104A discloses a supply chain security management method that obtains operational information from all nodes (requirements, design, coding, testing, etc.) in the software lifecycle and uses centralized "system management information" to perform vulnerability analysis on the code and documents output by each node to identify target nodes with security risks and issue warnings. This solution enhances the monitoring capability of the security status of the supply chain nodes themselves to a certain extent.

[0004] However, existing supply chain node management methods, including the aforementioned patents, mainly focus on the detection of single-point security flaws, that is, conducting independent vulnerability analysis on each node, detached from upstream and downstream connections, without fundamentally solving the problem of efficient and reliable collaboration among multiple nodes.

[0005] Specifically, this manifests as follows:

[0006] 1. Independent analysis of single nodes leads to rigid collaboration mechanisms: Conventional methods only focus on whether each node has defects, lacking analysis of the interaction between nodes. The adjustment of plans and handling of anomalies between nodes heavily rely on the central node or manual negotiation, and cannot make a fast and automated collaborative response to real-time linkage changes between multiple nodes (such as the cascading impact of a sudden drop in upstream node inventory on downstream production cycle).

[0007] 2. Lack of forward-looking collaboration capabilities: It can only provide post-event warnings based on vulnerabilities that have already occurred, and cannot use node prediction data to conduct forward-looking analysis of future risks, leaving the supply chain in a state of passive response.

[0008] Therefore, how to overcome the limitations of independent analysis of a single node and achieve reliable perception and automated collaborative response of the interaction status between multiple nodes is a technical problem that urgently needs to be solved in the field of supply chain management. Summary of the Invention

[0009] The purpose of this invention is to solve practical problems. It provides a method for efficient collaborative management and control of supply chain nodes based on blockchain analysis, which realizes reliable perception of the interaction status between multiple nodes, dual-path parallel risk judgment, fine-grained handling of multi-source anomalies, and collaborative integration of instructions.

[0010] The objective of this invention can be achieved through the following technical solution: a method for efficient collaborative management and control of supply chain nodes based on blockchain analysis, comprising the following steps:

[0011] S1. Network Construction and Interaction Definition: Construct a blockchain network and predefine the interaction dependencies between nodes in the target supply chain, as well as the disturbance judgment rules and deviation comparison rules, through smart contracts or on-chain configuration tables.

[0012] S2, Node Data Collection, Prediction and On-Chain: Collect the running status data of each node in real time, generate prediction data for future periods based on this data, and upload the running status data and prediction data to the blockchain network after signing.

[0013] S3, Dual-Path Linkage Judgment and Instruction Generation: The interaction analysis program deployed on the blockchain network synchronously executes the following two judgment branches based on the interaction dependencies and the data uploaded by each node:

[0014] Chain effect assessment and judgment branch (risk prediction of associated nodes triggered by node anomalies): All nodes are independently detected according to the disturbance judgment rules, the disturbance source node is identified, and the associated nodes are bidirectionally traversed along the resource dependency direction in the interaction dependency relationship for each disturbance source node. The disturbance diffusion coefficient of the associated nodes is calculated and weighted and merged according to the propagation parameters in the interaction dependency relationship, the disturbed node is identified, and the disturbance influence domain composed of the disturbance source node and the disturbed node is determined. A comprehensive risk pre-adjustment instruction is generated for the nodes in the domain.

[0015] Forward-looking plan deviation determination branch (node ​​status prediction and future plan pre-adjustment analysis): Based on the deviation comparison rules, the prediction data of each node is compared with the global collaborative framework to obtain the prediction deviation value of each node. Then, based on the interaction dependency relationship, the interaction impact analysis of the prediction deviation value of each node is performed to obtain the comprehensive prediction deviation value of each node affected by the deviation of related nodes. Based on this, plan pre-adjustment instructions are generated for the relevant nodes.

[0016] S4. Collaborative Execution Steps: Issue comprehensive risk pre-adjustment instructions and plan pre-adjustment instructions to the corresponding target node execution systems to drive them to adjust their local business parameters.

[0017] Furthermore, the interaction dependencies in S1 are stored in a directed graph structure. Each node in the graph corresponds to a node identifier. Each directed edge contains a source node identifier, a target node identifier, a resource type identifier, and propagation parameters. The propagation parameters include at least one of the following: attenuation factor, interaction influence weight, and risk threshold. The resource type identifier is used to distinguish between physical materials, information flow, or capital flow.

[0018] Furthermore, the process of generating prediction data based on operational status data in S2 includes:

[0019] Collect operational status data for this node, including one or more of the following: inventory change events, production progress events, equipment status events, and order delivery events;

[0020] Predictive data is obtained from a prediction model deployed locally on each node. The prediction model is a time series prediction model or a machine learning prediction model. Its input is a time series of historical operating status data of the local node, and its output is the predicted resource quantity or predicted job completion time for at least one time slot in the future period. The prediction model periodically uses the latest collected operating status data for incremental training or parameter calibration. The operating status data and prediction data are uploaded to the blockchain in the form of hash digests, and the original data is stored in the local database of the node.

[0021] Furthermore, in the chain effect evaluation branch, the method for determining the disturbance source node is as follows:

[0022] Based on the disturbance determination rules, the running status data of each node is preprocessed to obtain a set of quantitative indicators. Each quantitative indicator in the set is checked one by one, and its deviation from its own standard baseline is calculated. When the deviation of at least one quantitative indicator exceeds the preset deviation safety threshold, the node is determined to be a disturbance source node, and all quantitative indicators with deviation exceeding the safety threshold are marked as abnormal indicators.

[0023] Furthermore, the method for obtaining the disturbance influence domain is as follows:

[0024] T1. Comprehensive Disturbance Intensity Calculation: The initial disturbance intensity value of each disturbance source node is determined based on the deviation degree of the abnormal indicators and the preset abnormality type mapping table. The comprehensive disturbance intensity of the node is obtained by comprehensively calculating the initial disturbance intensity values ​​of all abnormal indicators of the disturbance source node. The comprehensive disturbance intensity is a normalized value used to characterize the overall abnormality severity of the node.

[0025] T2. Disturbance direction identification: Based on the anomaly type with the highest preset importance or the greatest deviation among all anomaly indicators, determine the disturbance direction identification of the disturbance source node, including downstream impact type, upstream impact type, or bidirectional impact type.

[0026] T3. Single-source disturbance diffusion calculation: For each disturbance source node, according to the disturbance direction identifier, traverse the associated nodes along the corresponding resource dependency direction, and calculate the disturbance diffusion coefficient according to the linear decay model. Specifically, obtain the directed path length d from the disturbance source node i to the associated node j, read the decay factor from the interaction dependency relationship and mark it as α, and calculate the disturbance diffusion coefficient Rj using the linear decay model: Rj=Ei×(1-α×d), where Ei is the comprehensive disturbance intensity value, α is ×d≤1, otherwise Rj=0 (indicating decay to no effect).

[0027] T4. Multi-source disturbance superposition: For an associated node that has received one or more disturbance diffusion coefficients, the disturbances are combined according to a preset superposition rule to obtain the comprehensive disturbance diffusion coefficient of the associated node. The superposition rule includes one of taking the maximum value, summing, or weighted summing.

[0028] T5. Disturbed Node Determination: When the comprehensive disturbance diffusion coefficient of a certain associated node is greater than the preset risk threshold corresponding to that node, the node is marked as a disturbed node and included in the disturbance influence domain together with the disturbance source node.

[0029] Furthermore, the method for generating the comprehensive risk pre-adjustment instruction includes:

[0030] Construct a pre-adjustment table: Sort all disturbance source nodes in descending order of comprehensive disturbance intensity to generate a pre-adjustment table of disturbance source nodes; Sort all disturbed nodes in descending order of comprehensive disturbance diffusion coefficient to generate a pre-adjustment table of disturbed nodes.

[0031] Anomaly source instruction generation: According to the pre-adjustment table, the preset anomaly repair action template is matched according to all anomaly index types and node type identifiers of each disturbance source node. If there are multiple anomaly index types, a combined instruction containing multiple actions is generated or multiple atomic instructions are generated in sequence according to priority. The action strength and priority are determined according to the comprehensive disturbance strength, high-priority atomic instructions are generated, and the generated instructions are recorded as subsequent resource conflict constraints.

[0032] Anomaly association instruction generation: According to the order of the pre-adjustment table, based on the anomaly source type, node type identifier and comprehensive disturbance diffusion coefficient of each disturbed node, the preset risk avoidance action template is matched, the action intensity is determined according to the comprehensive disturbance diffusion coefficient, and an atomic instruction is generated, which has a lower priority than the anomaly source instruction.

[0033] Instruction merging and arbitration: Merge two types of atomic instructions. When multiple instructions compete for the same resource, arbitration is carried out according to the rule that the instruction from the abnormal source takes precedence over the instruction from the related point, and the instruction with higher strength / coefficient among the same type takes precedence. The arbitrated instructions are then aggregated into the same data packet to form the comprehensive risk pre-adjustment instruction.

[0034] Furthermore, the specific process of obtaining the comprehensive prediction bias value through interactive analysis in the prospective bias branch includes:

[0035] The predicted data of each node is compared with the expected value of the corresponding time slot in the global collaboration framework. The global collaboration framework is stored in the form of time series, which includes the expected amount of resources or the expected job completion time of each node in multiple future time slots, and the prediction deviation value of that node is obtained.

[0036] Based on the interaction dependency, the interaction influence weights between nodes are determined. For each target node, its own prediction deviation value is added to the prediction deviation values ​​of all associated nodes according to their respective interaction influence weights to obtain the comprehensive prediction deviation value of the node.

[0037] Furthermore, in step S4, before issuing the comprehensive risk pre-adjustment instruction and the plan pre-adjustment instruction, a step of instruction coordination and fusion is also included:

[0038] The comprehensive forecast deviation value in the planning adjustment instructions generated by the forward-looking deviation branch is used as an additional influencing factor to correct the comprehensive risk adjustment instructions generated by the cascading effect assessment branch.

[0039] When the overall prediction deviation value is consistent with the sign direction of the disturbance diffusion coefficient, the corresponding pre-adjustment action is strengthened;

[0040] When the sign of the comprehensive prediction deviation value is opposite to that of the disturbance diffusion coefficient, arbitration is carried out according to the preset priority rules, in which the priority of the disturbance diffusion coefficient is higher than that of the comprehensive prediction deviation value.

[0041] Compared with the prior art, the advantages of this invention are:

[0042] 1. This invention achieves decentralized real-time linkage response through "blockchain network construction + interaction dependency analysis + two-way chain effect assessment". Specifically, it constructs a consortium blockchain network, pre-stores the interaction dependencies between nodes in the form of a directed graph, and when a node detects an anomaly, it traverses upstream and downstream related nodes bidirectionally along the resource dependency direction. Based on the linear decay model, it independently calculates the propagation impact of each anomaly source and performs multi-source disturbance superposition to determine the disturbed node. It then determines the disturbance impact domain containing the disturbance source node and the anomaly related node, and generates deterministic and automatically executable control decision instructions for multiple nodes within the domain. This effectively solves the multi-source risk management problem in complex supply chains and accurately handles multiple concurrent anomaly source scenarios.

[0043] 2. This invention also achieves predictive collaboration of future risks through "forward-looking deviation interaction analysis + plan correction". Specifically, when monitoring nodes in real time, future predictions are made simultaneously. The predicted data is compared with the global collaboration framework to obtain the prediction deviation value of a single node. The prediction deviation value of each node is accumulated and interactively analyzed to obtain the comprehensive prediction deviation value of each node affected by the deviation of related nodes. Based on this, a plan adjustment instruction is generated. Each node can perceive the chain effect of upstream and downstream prediction deviations and make capacity adjustments or inventory transfers in advance. With the global collaboration framework as the benchmark, the entire chain is driven to converge toward the optimal plan, reducing the bullwhip effect. The prediction data is uploaded to the chain with a hash digest to protect trade secrets. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the overall method of the present invention;

[0045] Figure 2 This is a flowchart of the instruction coordination fusion and execution method in Embodiment 2 of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0047] Example 1:

[0048] This invention discloses a method for efficient collaborative management and control of supply chain nodes based on blockchain analysis. Please refer to [link / reference]. Figure 1 It includes the following steps:

[0049] S1. Network Construction and Interaction Definition: Construct a blockchain network and predefine the interaction dependencies between nodes in the target supply chain, as well as the disturbance judgment rules and deviation comparison rules, through smart contracts or on-chain configuration tables. The interaction dependencies are stored in a directed graph structure, with each node in the graph corresponding to a node identifier. Each directed edge contains a source node identifier, a target node identifier, a resource type identifier, and propagation parameters. The propagation parameters include at least one of the following: attenuation factor, interaction influence weight, and risk threshold. The resource type identifier is used to distinguish between physical materials, information flow, or capital flow.

[0050] Build a decentralized trust infrastructure and predefine the interaction dependencies between nodes to provide basic data and rules for subsequent analysis;

[0051] S2, Node Data Acquisition, Prediction, and On-Chain: Real-time acquisition of operational status data from each node, generating prediction data for future periods based on this data, and uploading the operational status data and prediction data to the consortium blockchain network after signing. Operational status data includes one or more of the following: inventory change events, production progress events, equipment status events, and order delivery events. Prediction data is obtained from prediction models deployed locally on each node. Hash digests are calculated for both operational status data and prediction data, and the digests and timestamps are signed and broadcast to the blockchain. The original data is stored in the node's local database, and the hash digests are used for subsequent verification and auditing.

[0052] Here, the prediction model is a time series prediction model or a machine learning prediction model. Its input is the time series of the historical running status data of this node, and its output is the predicted resource amount or the predicted job completion time for at least one time slot in the future period. The prediction model periodically uses the latest collected running status data for incremental training or parameter calibration. The running status data and prediction data are uploaded to the chain in the form of hash digests, and the original data is stored in the node's local database.

[0053] S3, Dual-Path Linkage Judgment and Instruction Generation: Automatically triggered by the interactive analysis program deployed on the blockchain network, based on the interaction dependency, the data uploaded by each node, and the disturbance judgment rules and deviation comparison rules, the following two judgment branches are executed simultaneously, and each generates a control instruction:

[0054] Chain effect assessment and judgment branch: Based on the disturbance judgment rules, all nodes are independently detected to identify the disturbance source node. The single node anomaly serves as a signal source to trigger the collaborative analysis of all nodes with which it has an interaction relationship. That is, for each disturbance source node, the associated nodes are bidirectionally traversed along the resource dependency direction in the interaction dependency relationship. The disturbance diffusion coefficient of the associated nodes is calculated based on the propagation parameters in the interaction dependency relationship and weighted and merged to identify the disturbed node. The disturbance influence domain composed of the disturbance source node and the disturbed node is determined, and a comprehensive risk pre-adjustment instruction is generated for the nodes in the domain.

[0055] Forward-looking plan deviation determination branch: Based on the deviation comparison rules, the prediction data of each node is compared with the global collaborative framework to obtain the prediction deviation value of each node. Then, based on the interaction dependency relationship, the interaction impact analysis of the prediction deviation value of each node is performed to obtain the comprehensive prediction deviation value of each node affected by the deviation of related nodes. Based on this, plan pre-adjustment instructions are generated for the relevant nodes.

[0056] S4. Collaborative Execution Steps: Issue comprehensive risk pre-adjustment instructions and plan pre-adjustment instructions to the corresponding target node execution systems to drive them to adjust their local business parameters.

[0057] In the chain reaction assessment branch of S3, the method for determining the disturbance source node is as follows:

[0058] Based on the disturbance judgment rules, the running status data of each node is preprocessed to obtain a set of quantitative indicators. Each quantitative indicator in the set is checked one by one, and its deviation from its own standard baseline is calculated. When the deviation of at least one quantitative indicator exceeds the preset deviation safety threshold, the node is judged as the disturbance source node, and all quantitative indicators with deviation exceeding the safety threshold are marked as abnormal indicators.

[0059] The method for obtaining the disturbance influence domain is as follows:

[0060] T1. Calculation of Comprehensive Disturbance Intensity: The initial disturbance intensity value of each disturbance source node is determined based on the deviation of the abnormal indicators and the preset abnormality type mapping table. The comprehensive disturbance intensity of the node is calculated by combining the initial disturbance intensity values ​​of all abnormal indicators of the disturbance source node. The comprehensive disturbance intensity is a normalized value used to characterize the overall abnormality severity of the node.

[0061] T2. Disturbance direction identification: Based on the anomaly type with the highest preset importance or the greatest deviation among all anomaly indicators, determine the disturbance direction identification of the disturbance source node, including downstream impact type, upstream impact type, or bidirectional impact type.

[0062] T3. Single-source disturbance diffusion calculation: For each disturbance source node, according to the disturbance direction identifier, traverse the associated nodes along the corresponding resource dependency direction to obtain the directed path length d from the disturbance source node i to the associated node j. Read the attenuation factor from the interaction dependency relationship and mark it as α. Use the linear attenuation model to calculate the disturbance diffusion coefficient Rj: Rj=Ei×(1-α×d), where Ei is the comprehensive disturbance intensity value, α is ×d≤1, otherwise Rj=0 (indicating attenuation to no effect);

[0063] Among them, when the disturbance direction is identified as downstream influence type, the disturbance diffusion coefficient is used as the positive disturbance diffusion coefficient; when the disturbance direction is identified as upstream influence type, the disturbance diffusion coefficient is used as the negative disturbance diffusion coefficient; when the disturbance direction is identified as bidirectional influence type, both the positive disturbance diffusion coefficient and the negative disturbance diffusion coefficient are obtained.

[0064] T4. Multi-source disturbance superposition: For an associated node that has received one or more disturbance diffusion coefficients, the disturbances are combined according to a preset superposition rule to obtain the comprehensive disturbance diffusion coefficient of the associated node. The superposition rule includes taking the maximum value, summing or weighted summing.

[0065] T5. Disturbed Node Determination: When the comprehensive disturbance diffusion coefficient (forward or reverse) of a certain associated node is greater than the preset risk threshold corresponding to that node, the node is marked as a disturbed node and included in the disturbance influence domain together with the disturbance source node.

[0066] The methods for generating comprehensive risk pre-adjustment instructions include:

[0067] (a) Constructing a pre-adjustment table

[0068] Sort all disturbance source nodes in descending order of comprehensive disturbance intensity to generate a disturbance source node pre-adjustment table;

[0069] Sort all disturbed nodes in descending order of comprehensive disturbance diffusion coefficient to generate a disturbed node pre-adjustment table;

[0070] (ii) Generation of exception source instructions

[0071] Perform the following operations on each disturbance source node in the order of the pre-tuning table:

[0072] Based on all abnormal indicator types and node type identifiers of each disturbance source node, a preset abnormal repair action template is matched. If there are multiple abnormal indicator types, a combined instruction containing multiple actions is generated or multiple atomic instructions are generated sequentially according to priority. The action intensity and priority are determined based on the comprehensive disturbance intensity, a high-priority atomic instruction is generated, and the generated instructions are recorded as subsequent resource conflict constraints.

[0073] (III) Generation of Abnormal Association Instructions

[0074] The order of receiving the pre-adjustment table of disturbed nodes is as follows: according to the anomaly source type, node type identifier and comprehensive disturbance diffusion coefficient of each disturbed node, the preset risk avoidance action template is matched, the action intensity is determined according to the comprehensive disturbance diffusion coefficient, and an atomic instruction is generated, which has a lower priority than the anomaly source instruction.

[0075] (iv) Command merging and conflict resolution

[0076] The atomic instructions of the disturbance source node and the atomic instructions of the disturbed node are merged. When multiple instructions compete for the same resource, arbitration is carried out according to the rule that the abnormal source instruction takes precedence over the related point instruction, and the one with higher strength / coefficient among the same type takes precedence. The arbitration results are aggregated into the same data packet to form a comprehensive risk pre-adjustment instruction.

[0077] In S3, the specific process of obtaining the comprehensive forecast bias value through interactive analysis in the forward bias branch includes:

[0078] The predicted data of each node is compared with the expected value of the corresponding time slot in the global collaboration framework. The global collaboration framework is stored in the form of time series, which includes the expected amount of resources or the expected job completion time of each node in multiple future time slots, to obtain the prediction deviation value of the node, including the deviation direction and deviation magnitude.

[0079] Based on the interaction dependency, the interaction influence weight between nodes is determined. The closer the upstream and downstream resource dependency, the greater the deviation influence weight. For each target node, its own prediction deviation value is added to the prediction deviation values ​​of all related nodes according to their respective interaction influence weights to obtain the comprehensive prediction deviation value of the node.

[0080] Example 2:

[0081] See Figure 2 In S4, before issuing the comprehensive risk pre-adjustment instruction and the plan pre-adjustment instruction, there is also an instruction coordination and execution step:

[0082] Read the comprehensive control instructions and plan revision instructions, and use the comprehensive forecast deviation value in the plan pre-adjustment instructions generated by the forward-looking deviation branch as an additional influencing factor to revise the comprehensive risk pre-adjustment instructions generated by the chain effect assessment branch;

[0083] Specifically, it includes:

[0084] Obtain the disturbance source nodes and abnormal associated nodes within the disturbance influence domain, and use them as target correction nodes;

[0085] For each target correction node, its comprehensive disturbance diffusion coefficient and comprehensive prediction deviation value are obtained. The comprehensive disturbance diffusion coefficient of the disturbance source node is preset to the highest value of 1.0.

[0086] When the sign direction of the comprehensive prediction deviation value is consistent with that of the comprehensive disturbance diffusion coefficient, the action type of the comprehensive risk pre-adjustment instruction remains unchanged, and the execution intensity of the instruction is increased linearly or in stages according to the magnitude of the comprehensive prediction deviation value, and a new instruction with intensity correction is output.

[0087] When the sign of the comprehensive prediction deviation value is opposite to that of the comprehensive disturbance diffusion coefficient, arbitration is carried out according to the preset priority rules, wherein the priority of the disturbance diffusion coefficient is higher than that of the comprehensive prediction deviation value.

[0088] For example, when the overall disturbance diffusion coefficient is greater than the maximum value of the preset rule range, the overall control command shall prevail; when the overall disturbance diffusion coefficient is greater than the minimum value of the preset rule range, the plan correction command shall prevail; when the overall disturbance diffusion coefficient is within the preset rule range, the new command after arbitration shall be output by merging the disturbance diffusion coefficient and the overall prediction deviation value according to the preset weights.

[0089] By constructing an anomaly source pre-adjustment table and a disturbed node pre-adjustment table, and sorting them in descending order of comprehensive disturbance intensity / risk coefficient; anomaly source instructions take precedence over related point instructions, and those with higher coefficients of the same type take priority in arbitrating resource competition; furthermore, the comprehensive prediction deviation value is used as an additional influencing factor, and is used to judge the consistency of direction and integrate with risk instructions in a hierarchical manner, so as to avoid execution chaos caused by instruction conflict and achieve the organic unity of current risk control and future plan adjustment.

[0090] In summary, by constructing a consortium blockchain network and pre-storing the interaction dependencies between nodes, a unified decentralized real-time linkage response and forward-looking collaboration is achieved. On the one hand, when a node is abnormal, the upstream and downstream nodes are traversed bidirectionally along the resource dependency direction. Based on the linear decay model, the propagation impact of each abnormal source is independently calculated and multi-source disturbances are superimposed to locate the disturbance impact domain. Deterministic and automatically executable control decision instructions are generated for nodes within the domain. This effectively solves the problem that existing technologies suffer from rigid collaboration mechanisms and an inability to respond quickly to real-time linkage changes of multiple nodes due to the lack of interaction analysis between nodes.

[0091] On the other hand, while monitoring in real time, future predictions are made simultaneously. The prediction data is compared with the global collaborative framework, and the prediction deviation values ​​of each node are accumulated and interactively analyzed to obtain the comprehensive prediction deviation value affected by the deviation of related nodes. Based on this, a pre-adjustment instruction is generated, enabling each node to perceive the chain effect of upstream and downstream prediction deviations and make capacity adjustments or inventory transfers in advance. This overcomes the shortcomings of traditional technologies that can only provide post-event warnings and cannot proactively analyze future risks, resulting in the supply chain always being in a passive response state.

[0092] It enables parallel perception and collaborative management of current risks and future deviations through dual paths. It not only automatically completes multi-source risk propagation calculation and resource arbitration within milliseconds, forming executable instructions with clear priorities and no conflicts, but also drives the entire chain to converge toward the optimal plan through interactive prediction, upgrading the supply chain from "passive firefighting" to "proactive prevention", significantly improving the collaborative response speed of multiple nodes and the overall operational efficiency.

[0093] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto; any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A method for efficient collaborative management and control of supply chain nodes based on blockchain analysis, characterized by: Includes the following steps: S1. Network Construction and Interaction Definition: Construct a blockchain network and predefine the interaction dependencies between nodes in the target supply chain, as well as the disturbance judgment rules and deviation comparison rules. S2, Node Data Collection, Prediction and On-Chain: Collect the running status data of each node in real time, generate prediction data for future periods based on this data, and upload the running status data and prediction data to the blockchain after signing them. S3. Dual-path linkage judgment and instruction generation: Based on the interaction dependency relationship and the data uploaded by each node, the following two judgment branches are executed synchronously: Chain effect assessment and judgment branch: Independently detect all nodes, identify disturbance source nodes, bidirectionally traverse associated nodes along the resource dependency direction in the interaction dependency relationship for each disturbance source node, calculate the disturbance diffusion coefficient of associated nodes based on the propagation parameters in the interaction dependency relationship and weighted merge, identify disturbed nodes, determine the disturbance influence domain formed by the merger of disturbance source nodes and disturbed nodes, and generate comprehensive risk pre-adjustment instructions for nodes in the domain. Forward-looking planning deviation determination branch: The forecast data of each node is compared with the global collaborative framework to obtain the forecast deviation value of each node. Then, based on the interaction dependency, the interaction impact analysis of the forecast deviation value of each node is carried out to obtain the comprehensive forecast deviation value of each node affected by the deviation of related nodes. Based on this, the plan pre-adjustment instructions are generated for the relevant nodes. S4. Collaborative Execution Steps: Issue comprehensive risk pre-adjustment instructions and plan pre-adjustment instructions to the corresponding target node execution systems to drive them to adjust their local business parameters.

2. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 1, characterized in that: The interaction dependencies in S1 are stored in a directed graph structure. Each node in the graph corresponds to a node identifier. Each directed edge contains a source node identifier, a target node identifier, a resource type identifier, and propagation parameters. The propagation parameters include at least one of the following: attenuation factor, interaction influence weight, and risk threshold.

3. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 2, characterized in that: The process of generating prediction data based on operational status data in S2 includes: The system collects operational status data for this node and obtains prediction data from prediction models deployed locally on each node. These prediction models are either time series prediction models or machine learning prediction models. Their input is a time series of historical operational status data for this node, and their output is the predicted resource quantity or predicted job completion time for at least one time slot in the future period.

4. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 3, characterized in that: In the chain reaction assessment branch, the method for determining the disturbance source node is as follows: Based on the disturbance determination rules, the running status data of each node is preprocessed to obtain a set of quantitative indicators. Each quantitative indicator in the set is checked one by one, and its deviation from its own standard baseline is calculated. When the deviation of at least one quantitative indicator exceeds the preset deviation safety threshold, the node is determined to be a disturbance source node, and all quantitative indicators with deviations exceeding the safety threshold are marked as abnormal indicators.

5. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 4, characterized in that: The disturbance influence domain is obtained as follows: T1. Comprehensive Disturbance Intensity Calculation: Determine the initial disturbance intensity value of each disturbance source node based on the deviation degree of the abnormal indicators and the preset abnormality type mapping table, and calculate the comprehensive disturbance intensity of the node by comprehensively calculating the initial disturbance intensity values ​​of all abnormal indicators. T2. Disturbance direction identification: Based on the anomaly type with the highest preset importance or the greatest deviation among all anomaly indicators, determine the disturbance direction identification of the disturbance source node, including downstream impact type, upstream impact type, or bidirectional impact type. T3. Single-source disturbance diffusion calculation: For each disturbance source node, according to the disturbance direction identifier, traverse the associated nodes along the corresponding resource dependency direction and calculate the disturbance diffusion coefficient according to the linear decay model. T4. Multi-source disturbance superposition: For an associated node that has received one or more disturbance diffusion coefficients, the disturbances are combined according to a preset superposition rule to obtain the comprehensive disturbance diffusion coefficient of the associated node. T5. Disturbed Node Determination: When the comprehensive disturbance diffusion coefficient of a certain associated node is greater than the preset risk threshold corresponding to that node, the node is marked as a disturbed node and included in the disturbance influence domain together with the disturbance source node.

6. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 5, characterized in that: The methods for generating the comprehensive risk pre-adjustment instruction include: Constructing a pre-adjustment table: Sort the disturbance source nodes in descending order of comprehensive disturbance intensity to generate a pre-adjustment table of disturbance source nodes, and sort the disturbed nodes in descending order of comprehensive disturbance diffusion coefficient to generate a pre-adjustment table of disturbed nodes; According to the order of the pre-adjustment table, the repair template is matched according to the abnormal index type and node type identifier of each disturbance source node, the action intensity and priority are determined according to the comprehensive disturbance intensity, and the abnormal source instruction is generated. According to the order of the pre-adjustment table, the avoidance template is matched based on the anomaly source type, node type identifier and comprehensive disturbance diffusion coefficient of each disturbed node, the action intensity is determined based on the comprehensive disturbance diffusion coefficient, and anomaly association instructions are generated. By merging the two types of instructions mentioned above, when multiple instructions compete for the same resource, arbitration is carried out according to the rules that the abnormal source instruction takes precedence over the related point instruction, and the instruction with higher strength / coefficient among the same type takes precedence. The arbitration results are then aggregated into the same data packet to form the comprehensive risk pre-adjustment instruction.

7. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 6, characterized in that: The specific process for obtaining the comprehensive forecast deviation value includes: Based on the interaction dependency, the interaction influence weights between nodes are determined. For each target node, its own prediction deviation value is added to the prediction deviation values ​​of all associated nodes according to their respective interaction influence weights to obtain the comprehensive prediction deviation value of the node.

8. The efficient collaborative management and control method for supply chain nodes based on blockchain analysis according to claim 7, characterized in that: In step S4, before issuing the comprehensive risk pre-adjustment instruction and the plan pre-adjustment instruction, there is also an instruction coordination and integration step: the comprehensive forecast deviation value in the plan pre-adjustment instruction is used as an additional influencing factor to correct the comprehensive risk pre-adjustment instruction generated by the chain effect assessment branch.

Images