Supply chain full-link real-time monitoring method, device and equipment based on digital twinning and medium of supply chain full-link real-time monitoring method and device

Abstract

The invention relates to the technical field of supply chain real-time monitoring. The supply chain full-link real-time monitoring method, device and equipment based on digital twinning and the medium thereof are provided, and the method comprises the following steps: carrying out real-time optimization processing on segmented sensing data streams to generate a local optimization instruction set; performing cross-segment conflict detection processing on the local optimization instruction set to generate a conflict identification signal; performing coordination optimization processing based on the conflict identification signal, and generating a strategy instruction after coordination; performing execution processing on the coordinated strategy instruction, and driving a physical execution terminal; carrying out aggregation processing on segmented index vectors output by the digital twin sub-model to generate a full-link operation situation map; and the full-link operation situation map is analyzed and processed based on a preset performance deviation threshold, and a dynamic recombination instruction is generated, so that the technical effects of enhancing the network topology adaptation capability, optimizing the multi-target coordination efficiency and improving the cascade effect quantization precision are achieved.

Description

Technical Field

[0001] This invention relates to the field of real-time supply chain monitoring technology, and in particular to a method, apparatus, equipment and medium for real-time monitoring of the entire supply chain based on digital twins. Background Technology

[0002] Real-time monitoring technology across the entire supply chain, as a core means to improve operational efficiency, has become a research hotspot in the fields of modern logistics and intelligent manufacturing. Its core objective is to achieve optimized resource allocation, rapid response to anomalies, and adaptive system adjustment through full-process visualized management of multiple links such as procurement, production, warehousing, and transportation.

[0003] Traditional technologies employ static segmented monitoring models for supply chain monitoring, but these models cannot dynamically adapt to changes in network topology, easily leading to cross-segment resource conflicts and delayed responses. The conflict coordination process relies on human experience to formulate scheduling strategies, which makes it difficult to achieve Pareto optimality when facing multi-objective optimization problems, resulting in poor computational efficiency. Furthermore, global situational awareness uses simple data aggregation methods, which cannot accurately quantify cascading effects, causing reorganization decisions to deviate from actual needs. Summary of the Invention

[0004] Therefore, it is necessary to provide methods, devices, equipment and media for real-time monitoring of the entire supply chain based on digital twins to address the above-mentioned technical problems, so as to achieve the technical effects of enhancing network topology adaptability, optimizing multi-objective coordination efficiency and improving the quantification accuracy of cascading effects.

[0005] Firstly, this application provides a method for real-time monitoring of the entire supply chain based on digital twins, the method comprising:

[0006] The physical supply chain network is analyzed for business attributes and geographical topology to obtain a set of logical segments; the set of logical segments is then modeled using digital twins to generate digital twin sub-models.

[0007] The segmented sensor data streams are collected by IoT terminals deployed in each logical segment; the segmented sensor data streams are optimized in real time to generate local optimization instruction sets.

[0008] Cross-segment conflict detection processing is performed on the local optimization instruction set to generate conflict identification signals; coordination optimization processing is performed based on the conflict identification signals to generate coordinated policy instructions; and the coordinated policy instructions are executed to drive the physical execution terminal.

[0009] The segmented indicator vectors output by the digital twin model are aggregated to generate a full-link operation status map; the full-link operation status map is analyzed and processed based on a preset performance deviation threshold to generate dynamic reassembly instructions.

[0010] In one embodiment, the entire link operation status diagram is analyzed and processed based on a preset performance deviation threshold to generate a dynamic reconfiguration instruction, including:

[0011] Perform multi-dimensional bottleneck localization processing on the end-to-end operational status map to identify key performance deviation areas;

[0012] Topological dependency analysis is performed on key performance deviation regions to generate segmented impact assessment results.

[0013] Adaptive reorganization decision processing is performed based on segmented impact assessment and preset performance deviation thresholds to generate dynamic reorganization instructions.

[0014] In one embodiment, topological dependency analysis is performed based on key performance deviation regions to generate segmented impact assessment results, including:

[0015] Multi-level dependency path extraction is performed on key performance deviation areas to generate topological propagation chains;

[0016] The following formula is used to quantify the cascading effects of the topological propagation chain and generate initial influence weights:

[0017]

[0018] in, Indicates the initial influence weight. Indicates the source deviation intensity. Indicates the first The criticality factor of each node This represents the industry experience attenuation coefficient. Indicates the distance from the source node to the... Topological distance between nodes, Indicates the criticality of the node's business. Indicates the node buffer capacity factor. Indicates the total number of nodes in the propagation path;

[0019] The initial influence weights are dynamically adjusted based on the real-time system status to generate the segmented influence evaluation results.

[0020] In one embodiment, the segmented indicator vectors output by the digital twin model are aggregated to generate a full-link operational status map, including:

[0021] The segmented indicator vectors are aligned in time and space to generate a time-synchronized indicator matrix.

[0022] The time synchronization index matrix is ​​subjected to topological weighted fusion processing to generate a node association strength map;

[0023] Multimodal situation rendering is performed based on node association strength maps to generate a full-link operational situation map.

[0024] In one embodiment, cross-segment conflict detection processing is performed on the local optimization instruction set to generate a conflict identification signal, including:

[0025] Perform instruction semantic parsing on the local optimization instruction set to extract segmented resource requirement features;

[0026] Based on the segmented resource demand characteristics, cross-segment dependency graph traversal processing is performed to identify conflict node pairs;

[0027] The conflict intensity of the conflicting node pairs is quantized to generate a conflict identification signal.

[0028] In one embodiment, the physical supply chain network is subjected to business attribute and geographic topology analysis to obtain a logical segment set, including:

[0029] Multi-dimensional business feature extraction is performed on the nodes of the physical supply chain network to generate business attribute clusters;

[0030] Spatial relationship modeling is performed on the nodes of the physical supply chain network to generate geographical topological partitions;

[0031] Adaptive fusion processing is performed on business attribute clustering and geographic topology partitioning to generate a logical segment set.

[0032] In one embodiment, coordination optimization processing is performed based on conflict identification signals to generate coordinated policy instructions, including:

[0033] Multi-agent negotiation modeling is performed on the conflict identification signal to generate a conflict resolution strategy space;

[0034] Based on the preset Pareto optimization criterion, the conflict resolution strategy space is subjected to non-dominated solution screening to generate a candidate strategy set.

[0035] The candidate policy set is dynamically prioritized and scheduled to generate coordinated policy instructions.

[0036] Secondly, this application also provides a real-time monitoring device for the entire supply chain based on digital twins, the device comprising:

[0037] The segmented modeling module is used to perform business attribute and geographic topology analysis on the physical supply chain network to obtain a set of logical segments; and to perform digital twin modeling on the set of logical segments to generate digital twin sub-models.

[0038] The segmented optimization module is used to collect segmented sensor data streams through IoT terminals deployed in each logical segment; perform real-time optimization processing on the segmented sensor data streams; and generate local optimization instruction sets.

[0039] The cross-segment coordination module is used to perform cross-segment conflict detection processing on the local optimization instruction set and generate conflict identification signals; perform coordination optimization processing based on the conflict identification signals to generate coordinated policy instructions; and perform execution processing on the coordinated policy instructions to drive the physical execution terminal.

[0040] The global monitoring module is used to aggregate the segmented indicator vectors output by the digital twin model to generate a full-link operation status map; based on the preset performance deviation threshold, it analyzes and processes the full-link operation status map to generate dynamic reorganization instructions.

[0041] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods in the first aspect of this application.

[0042] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the methods in the first aspect of this application.

[0043] The method, apparatus, equipment and media for real-time monitoring of the entire supply chain based on digital twins provided in this application form logical segments by dynamically analyzing the business attributes and geographical topology of the physical supply chain network, and constructing digital twin sub-models for each segment. This enables the models to better fit the actual structural changes of the supply chain network, thereby enhancing the network topology adaptability.

[0044] After generating local optimization instructions, potential conflicts are identified through cross-segment conflict detection, and coordination optimization is performed based on conflict identification signals. The above process can effectively handle the scheduling needs of multiple stages and multiple objectives, thereby optimizing the coordination efficiency of multiple objectives.

[0045] By aggregating the segmented index vectors output by the digital twin model to generate a full-link operational status map, and then combining it with a preset performance deviation threshold for analysis and generating dynamic reorganization instructions, the propagation impact of deviations in each link can be captured more accurately, thereby improving the quantification accuracy of cascading effects. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1This is a flowchart of a digital twin-based real-time monitoring method for the entire supply chain in one embodiment of the present invention.

[0048] Figure 2 This is a flowchart illustrating the aggregation of segmented indicator vectors output by a digital twin model to generate a full-link operational status diagram, as described in one embodiment of the present invention.

[0049] Figure 3 This is a structural diagram of a digital twin-based real-time monitoring device for the entire supply chain in one embodiment of the present invention. Detailed Implementation

[0050] To make the above-mentioned objects, features, and advantages of this application more apparent and understandable, the specific embodiments of this application will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0051] First, the application scenarios of the embodiments of this application are described. In the embodiments of this application, a method, apparatus, equipment, and medium for real-time monitoring of the entire supply chain based on digital twins are provided, applicable to scenarios such as end-to-end supply chain operation management in manufacturing, multi-stage collaborative supervision of integrated logistics enterprises, and full-process tracking of goods across regional supply chains.

[0052] The digital twin-based real-time monitoring method, device, equipment and medium for the entire supply chain based on digital twins provided in this application embodiment can also be applied to other application scenarios such as supply chain management and control of goods distribution in the new retail industry, multi-node collaborative monitoring of cross-border trade supply chains, and dynamic management of upstream and downstream supply chains in the intelligent manufacturing industry. This is only an example and does not limit the specific application scenarios.

[0053] like Figure 1 As shown, this application provides a method for real-time monitoring of the entire supply chain based on digital twins, the method including:

[0054] S101: Perform business attribute and geographical topology analysis on the physical supply chain network to obtain a set of logical segments; perform digital twin modeling on the set of logical segments to generate a digital twin sub-model.

[0055] For example, business attribute analysis is performed on the nodes of the physical supply chain network to extract business function types, process relationships and resource demand characteristics. At the same time, geographical topology analysis is performed on the nodes to sort out the spatial distribution pattern and path connectivity. The results of the two types of analysis are synergistically integrated to generate a logical segment set with business homogeneity and geographical relevance.

[0056] For each logical segment in the logical segment set, its physical entity composition and business operation rules are clarified, a digital mapping framework is constructed, static attribute data and dynamic operation data are integrated, a real-time mapping relationship between physical quantities and digital quantities is established, the digital replication and simulation capabilities are built, and a digital twin sub-model is generated.

[0057] S102: Collect segmented sensor data streams through IoT terminals deployed in each logical segment; perform real-time optimization processing on the segmented sensor data streams to generate local optimization instruction sets.

[0058] For example, adaptable IoT terminals are deployed for the business scenarios and sensing requirements of each logical segment to capture the status parameters, operating indicators, and environmental sensing data of physical entities in real time, and continuously transmit and integrate them to generate structured segmented sensing data streams. The segmented sensing data streams are cleaned, denoised, and standardized. Combined with the business operation rules and preset performance standards of each logical segment, real-time simulation analysis and optimization deduction are performed through digital twin sub-models to clarify the execution requirements for resource allocation, process adjustment, and status correction, and generate local optimization instruction sets.

[0059] S103: Perform cross-segment conflict detection processing on the local optimization instruction set to generate conflict identification signals; perform coordination optimization processing based on the conflict identification signals to generate coordinated policy instructions; perform execution processing on the coordinated policy instructions to drive the physical execution terminal.

[0060] For example, the resource demand characteristics and execution constraints of the local optimization instruction set are analyzed, and the cross-segment dependency relationship of the supply chain network is traversed and compared to identify resource competition, timing conflicts and process connection contradictions between logical segments, quantify the intensity and scope of the conflict, and generate conflict identification signals.

[0061] Based on the conflict identification signal analysis, conflict types, involved logical segments, and core factors, a conflict resolution strategy space is constructed by combining multi-objective optimization criteria. Candidate strategies are selected and optimized according to business priority and real-time operating status, generating coordinated strategy instructions. The execution path and operation standards of the coordinated strategy instructions are defined and sent to the corresponding physical execution terminals via a dedicated transmission interface to guide them in completing resource scheduling and process adjustments, thus driving the physical execution terminals.

[0062] S104: Aggregate the segmented index vectors output by the digital twin model to generate a full-link operation status diagram; analyze and process the full-link operation status diagram based on a preset performance deviation threshold to generate dynamic reconfiguration instructions.

[0063] For example, the segmented indicator vectors output by the digital twin model are aligned in time and space, and then weighted and fused according to the strength of the supply chain network topology association. The operational status information and association features of each logical segment are integrated to generate a full-link operational status map. Based on a preset performance deviation threshold, the key performance indicators and associations in the full-link operational status map are analyzed to locate performance deviation areas and assess the scope of impact. Combined with the adaptation requirements of the supply chain network, the adjustment direction and execution plan are determined, and dynamic reorganization instructions are generated after analysis and processing.

[0064] One embodiment of this application provides a real-time monitoring method for the entire supply chain based on digital twins. By dynamically analyzing the business attributes and geographical topology of the physical supply chain network to form logical segments, and constructing digital twin sub-models for each segment, the model can better fit the actual structural changes of the supply chain network, thereby enhancing the network topology adaptability.

[0065] After generating local optimization instructions, potential conflicts are identified through cross-segment conflict detection, and coordination optimization is performed based on conflict identification signals. The above process can effectively handle the scheduling needs of multiple stages and multiple objectives, thereby optimizing the coordination efficiency of multiple objectives.

[0066] By aggregating the segmented index vectors output by the digital twin model to generate a full-link operational status map, and then combining it with a preset performance deviation threshold for analysis and generating dynamic reorganization instructions, the propagation impact of deviations in each link can be captured more accurately, thereby improving the quantification accuracy of cascading effects.

[0067] In one embodiment, the entire link operation status diagram is analyzed and processed based on a preset performance deviation threshold to generate a dynamic reconfiguration instruction, including:

[0068] (1) Perform multi-dimensional bottleneck location processing on the full-link operation status map to identify key performance deviation areas.

[0069] For example, the core evaluation dimensions for multi-dimensional bottleneck positioning in the full-link operation status diagram are clearly defined, focusing on the core aspects of resource utilization, such as the rationality of resource allocation and resource turnover efficiency; the core aspects of process efficiency, such as the timeliness of node processing and the smoothness of cross-segment connection; and the core aspects of collaboration and adaptation, such as the accuracy of supply and demand matching and the fit of strategy execution.

[0070] Based on core indicators across various dimensions, key operational metrics covering all stages of the supply chain are selected, and the scope and statistical standards for metric data extraction are clearly defined. The actual performance data of the extracted key operational metrics are systematically compared one-way with industry-standard benchmarks and pre-defined supply chain operational specifications, marking abnormal metrics that exceed reasonable fluctuation ranges. Through a metric traceability mechanism, the specific supply chain stages, related nodes, and data sources corresponding to abnormal metrics are tracked, eliminating false anomalies that may be caused by data errors. Considering the impact weight, diffusion range, and degree of constraint on the overall efficiency of the entire supply chain, the key performance deviation areas are initially identified.

[0071] (2) Perform topological dependency analysis based on the key performance deviation area to generate segmented impact assessment results.

[0072] For example, using the critical performance deviation area as the initial analysis node, the system identifies logical segments in the supply chain network that have direct business interactions and resource flow connections with the critical performance deviation area, clarifying the type and strength of direct dependencies. It further traces indirect logical segments that have business transmission and process connection with the directly dependent logical segments, clarifying the indirect paths of deviation propagation. Finally, it extracts multi-level topological dependency links containing both direct and indirect dependencies, and analyzes the connection order and interaction rules of each logical segment within these links.

[0073] This analysis examines the characteristics of deviations in critical performance areas, assessing their propagation patterns and diffusion characteristics within topology-dependent links. Propagation methods include forward transmission, reverse feedback, and cross-influence. The analysis integrates core influencing factors such as business criticality, buffer capacity, and topology location importance for each logical segment. Business criticality is reflected in the proportion of core business load, buffer capacity in resource redundancy and emergency allocation space, and topology location importance in link hub attributes and irreplaceability. Based on these influencing factors, a quantitative evaluation framework is constructed, determining the weighting rules for each factor. The depth, breadth, and duration of the impact of deviations on each logical segment are systematically quantified, gradually generating segment impact assessment results reflecting the affected states of each logical segment.

[0074] (3) Based on the segmented impact assessment and the preset performance deviation threshold, adaptive reorganization decision processing is performed to generate dynamic reorganization instructions.

[0075] For example, the impact level data of each logical segment in the segmented impact assessment results are compared one by one with the preset performance deviation thresholds to classify logical segments into three categories: those whose impact level does not exceed the threshold, those whose impact level is close to the threshold, and those whose impact level exceeds the threshold. Logical segments whose impact level does not exceed the threshold can maintain normal operation, logical segments whose impact level is close to the threshold need to be closely monitored, and logical segments whose impact level exceeds the threshold need to be considered for reorganization and adjustment. This initially clarifies the core scope of the supply chain network that needs to be reorganized. Based on the reorganization scope, the core objectives of adaptive reorganization are determined. These core objectives include restoring the operational efficiency of the entire link, optimizing the adaptability of the topology structure, and improving the ability to resist risks. At the same time, the constraints in the reorganization process are identified, including resource supply limits, business continuity requirements, and cost control standards.

[0076] Based on the scope, core objectives, and constraints of the restructuring, and considering the topological characteristics of the supply chain network, business operation requirements, and current resource allocation, a set of multi-scenario restructuring solutions is designed, encompassing topology adjustments, resource reallocation, process optimization and restructuring, and node function adaptation. A comprehensive evaluation system is established to assess and rank the restructuring solutions in the set based on their technical feasibility, adaptability to business needs, expected optimization effects, and controllable potential risks. The relatively optimal restructuring solution is selected, and its specific adjustments are translated into actionable operational requirements, parameter configuration standards, and process change specifications, gradually generating dynamic restructuring instructions.

[0077] In one embodiment, topological dependency analysis is performed based on key performance deviation regions to generate segmented impact assessment results, including:

[0078] (1) Perform multi-level dependency path extraction processing on the key performance deviation area to generate a topology propagation chain.

[0079] For example, the core nodes and boundary ranges of the critical performance deviation area are clearly defined, and the business interaction objects and resource input / output relationships of the core nodes are identified. Based on business interaction and resource flow rules, directly dependent nodes that have direct data transmission, process connection, and resource supply relationships with the critical performance deviation area are identified, and the type, function, and connection strength of the directly dependent nodes are recorded. Starting from the directly dependent nodes, indirect dependent nodes that have indirect business transmission, cross-process collaboration, and related resource allocation are traced layer by layer to determine the level of indirect dependence and the association logic of nodes at each level. The directly dependent nodes and indirect dependent nodes at each level are integrated, and the nodes are arranged according to the order of dependency relationships and transmission direction, clarifying the connection paths and interaction standards between nodes, and generating a complete topology propagation chain.

[0080] The multi-level dependency path extraction process includes clarifying the core nodes and boundaries of key performance deviation areas, identifying direct dependency nodes and recording connection strength, tracing the hierarchical structure of indirect dependency nodes and determining the association logic, arranging the node sequence and clarifying the connection path, and integrating the topology propagation chain.

[0081] (2) Use the following formula to quantify the cascading effects of the topological propagation chain and generate the initial influence weights:

[0082]

[0083] in, Indicates the initial influence weight. Indicates the source deviation intensity. Indicates the first The criticality factor of each node This represents the industry experience attenuation coefficient. Indicates the distance from the source node to the... Topological distance between nodes, Indicates the criticality of the node's business. Indicates the node buffer capacity factor. This indicates the total number of nodes along the propagation path.

[0084] For example, based on the starting point characteristics of the topology propagation chain, and combined with the deviation type, duration, and initial impact range of the critical performance deviation region, the intensity of the source deviation is assessed. For each node in the topology propagation chain, the criticality factor of the node is determined according to its business capacity in the supply chain, functional irreplaceability, and supporting role in the entire chain. Referring to the deviation propagation patterns of similar scenarios in the industry and historical data statistics, an industry-experienced attenuation coefficient is determined. The path length from the source node to each node in the topology propagation chain and the number of connected nodes are calculated to obtain the topological distance of each node.

[0085] The core business proportion of each node and its impact on the overall process are analyzed to determine the node's business criticality. The resource redundancy and emergency adjustment space of each node are assessed to determine the node's buffer capacity factor. Following the formula logic, the correlation results of source deviation intensity, criticality factor of each node, topological distance and industry experience attenuation coefficient, and the relationship between node business criticality and buffer capacity factor are systematically integrated and calculated to generate initial impact weights.

[0086] The cascading effect quantification process includes source deviation intensity assessment, node criticality factor determination, industry experience attenuation coefficient reference, node topology distance calculation, node business criticality and buffer capacity factor analysis, and initial impact weight integration calculation.

[0087] (3) The initial influence weights are dynamically corrected based on the real-time system status to generate the segmented influence evaluation results.

[0088] For example, information such as device operating status, real-time resource inventory, order priority changes, and process execution progress of nodes involved in the topology propagation chain in a real-time system is collected. The correlation mechanism between the real-time system status and the initial impact weight is analyzed. For example, abnormal device operation may amplify the impact on nodes, sufficient resource inventory may reduce the impact on nodes, and high-priority orders may change the speed of impact propagation of nodes.

[0089] By combining the correlation mechanism, dynamic correction rules are formulated to clarify the adjustment direction and magnitude of the initial impact weights under different real-time states. Based on the dynamic correction rules, the initial impact weights are adjusted node-by-node, and the corrected results must comprehensively reflect the actual impact level of each node under the real-time system state. The correction results of all nodes are integrated to form an overall impact assessment covering the logical segments of the topology propagation chain, generating segmented impact assessment results.

[0090] The dynamic correction process includes real-time system status information collection, analysis of the correlation mechanism between status and initial weights, formulation of dynamic correction rules, node-by-node adjustment of initial impact weights, integration of correction results, and generation of segmented impact assessment results.

[0091] like Figure 2 As shown, the segmented indicator vectors output by the digital twin model are aggregated to generate a full-link operational status map, including:

[0092] S201: Perform spatiotemporal dimension alignment processing on the segmented index vectors to generate a time-synchronized index matrix.

[0093] For example, the timestamp information and spatial node identifiers contained in the segmented index vectors output by each logical segment are clearly defined, and a unified time reference (i.e., standard time scale) and spatial coordinate system (i.e., node topology numbering rules) are determined. For the segmented index vectors of different logical segments, the acquisition time of each index data is calibrated according to the time reference to eliminate time differences caused by data transmission delays in the time dimension.

[0094] The physical nodes and logical segment positions corresponding to each segmented indicator vector are matched according to the spatial coordinate system to ensure that the indicator data in the spatial dimension corresponds one-to-one with the actual nodes. The segmented indicator vectors that have undergone spatiotemporal calibration are integrated, and the indicator data are arranged in the order of time series and spatial nodes to generate a well-structured time-synchronized indicator matrix.

[0095] The spatiotemporal alignment process includes determining the time reference and spatial coordinate system, timestamp calibration, spatial node matching, and integration and arrangement of indicator data.

[0096] S202: Perform topological weighted fusion processing on the time synchronization index matrix to generate a node association strength map.

[0097] For example, the topological structure of the supply chain network is analyzed to clarify the business dependencies between nodes, including upstream and downstream collaboration, resource supply and demand, data interaction frequency, and impact transmission paths. Based on the topological structure and historical operational data, the association weights between nodes are determined; higher business dependencies and more frequent data interactions result in higher weights. Using a time-synchronized indicator matrix as a foundation, the indicator data of corresponding nodes in the matrix are weighted and calculated in conjunction with the association weights, thus integrating the mutual influence information between nodes. The weighted and integrated node association information is then presented graphically, using line thickness and color intensity to represent the association strength between nodes, generating a node association strength map.

[0098] The topology weighted fusion processing includes sorting out the topological structure relationships, determining the node association weights, calculating the weighted index data, and graphically presenting the association strength.

[0099] S203: Perform multimodal situation rendering based on node association strength map to generate a full-link operation situation map.

[0100] For example, the presentation format of multimodal situation rendering is determined, including but not limited to dynamic flowcharts showing changes in node operating status, heatmaps presenting the high and low distribution of indicator values, and correlation graphs strengthening the correlation strength between nodes. For node operating indicators in the node correlation strength graph, such as efficiency, load, and abnormal status, mapping rules are formulated between indicator values ​​and visual elements such as color gradients and icon sizes.

[0101] According to the mapping rules, the association strength information in the node association strength map is fused with the real-time operation indicators of each node and presented synchronously in a multimodal format. By integrating the rendering results of different modalities, the node status, association relationships, and overall operation trend of the entire link are clearly displayed in the same view, generating a full-link operation status map.

[0102] The multimodal situation rendering process includes determining the rendering presentation format, formulating mapping rules between indicators and visual elements, fusing rendering of correlation strength and operational indicators, and integrating multimodal results.

[0103] In one embodiment, cross-segment conflict detection processing is performed on the local optimization instruction set to generate a conflict identification signal, including:

[0104] (1) Perform instruction semantic parsing on the local optimization instruction set to extract segmented resource requirement features.

[0105] For example, the overall framework of the local optimization instruction set is outlined, clarifying the syntax structure, field meanings, and organization of each instruction. For each instruction, the semantics are parsed field by field to determine the logical segment identifier corresponding to the instruction, identify the resource type, specific specifications, and required quantity range to be invoked by the instruction, and clarify the time interval, execution order requirements, and insurmountable constraint boundaries of the resource requirements.

[0106] The parsed resource requirement information is validated to eliminate invalid information caused by semantic ambiguity or vague description, ensuring the completeness and accuracy of the requirement elements. The validated resource requirement information is then logically segmented and integrated, and the core characteristics and personalized requirements of the same logical segment in terms of resource requirement type, time distribution, priority ranking, etc., are extracted to generate the system's segmented resource requirement characteristics.

[0107] The instruction semantic parsing process includes instruction framework and structure analysis, instruction field semantic parsing, resource requirement element identification, requirement information verification, and segmented resource requirement feature extraction and integration.

[0108] (2) Based on the segmented resource demand characteristics, perform cross-segment dependency graph traversal processing to identify conflict node pairs.

[0109] For example, a pre-constructed cross-segment dependency graph of the supply chain network is retrieved to clarify the logical segment node information, resource flow paths between nodes, shared resource configuration, and interaction constraint rules contained in the graph. Based on the segment resource demand characteristics, traversal rules for the cross-segment dependency graph are formulated, clarifying the starting node, path order, and comparison dimensions.

[0110] Following the traversal rules, the process proceeds node by node along the links in the graph, comparing each node to determine if there are overlaps in resource demand categories, time intervals, total demand exceeding the available supply limit of the corresponding resource, and whether execution requirements contradict interaction constraint rules. Information on nodes with resource contention, time conflicts, or rule conflicts is recorded to further verify the authenticity and relevance of the conflicts, eliminate false conflicts, and identify conflicting node pairs with clear conflict relationships.

[0111] The cross-segment dependency graph traversal processing includes cross-segment dependency graph retrieval, traversal rule formulation, node association link traversal, cross-segment resource requirement comparison, conflict node information recording, conflict verification and misjudgment elimination, and conflict node pair determination.

[0112] (3) Perform conflict intensity quantization on the conflict node pairs to generate conflict identification signals.

[0113] For example, the importance level of conflict nodes to the types of resources involved in the entire supply chain is analyzed, and the scope of business links covered by the conflict, the degree of impact on related nodes, and the possible duration of the conflict are assessed. A multi-dimensional conflict intensity assessment system is constructed by combining the market scarcity of resources, the priority weight of corresponding businesses, and the ease of conflict resolution, and the measurement standards for each assessment dimension are clearly defined.

[0114] Based on the evaluation system, the degree of conflict between conflicting node pairs is comprehensively calculated to quantify the severity level of the conflict. Information such as the type of conflict, the identifiers of the logical segment nodes involved, the quantified intensity level, and the scope of conflict impact are integrated and transformed into a structured signal that conforms to the system's identification standards, generating a conflict identification signal.

[0115] The quantitative processing of conflict intensity includes the analysis of conflict impact factors, the construction of an assessment system, the clarification of measurement standards, the comprehensive calculation of conflict intensity, the conversion of structured signals, and the generation of conflict identification signals.

[0116] In one embodiment, the physical supply chain network is subjected to business attribute and geographic topology analysis to obtain a logical segment set, including:

[0117] (1) Perform multi-dimensional business feature extraction processing on the nodes of the physical supply chain network to generate business attribute clusters.

[0118] For example, the dimensions for extracting multi-dimensional business characteristics of nodes in the physical supply chain network are determined, including core aspects such as business type, functional positioning, resource demand characteristics, process association attributes, service target scope, and operating model. Specific information for each node under each extracted dimension is collected to clarify the node's business function boundaries, data interaction objects, and resource input / output specifications.

[0119] The collected business feature data is standardized to eliminate interference caused by differences in data representation between different nodes and ensure data comparability. Based on the similarity and correlation of business features of each node, a clustering algorithm is used to group and classify the nodes, so that nodes within the same group have similar business attribute features, generating business attribute clusters.

[0120] The multi-dimensional business feature extraction process includes determining the extraction dimensions, collecting node business information, standardizing feature data, clustering similarity data, and generating business attribute clusters.

[0121] (2) Perform spatial relationship modeling on the nodes of the physical supply chain network to generate geographical topology partitions.

[0122] For example, spatial elements such as geographic coordinates, actual geographical location distribution, surrounding transportation network layout, and geographical constraints of each node in the physical supply chain network are collected. Based on the collected spatial elements, a spatial relationship model between nodes is constructed to clarify the spatial proximity, path accessibility, and geographical region affiliation of the nodes.

[0123] Analyze the spatial correlation between each node and its surrounding nodes, and, considering factors such as geographical boundaries, transportation hub distribution, and regional economic characteristics, delineate preliminary geographical topological units. Verify the rationality of these preliminary geographical topological units, adjust units with ambiguous boundaries or loose connections, and ensure that nodes within each unit have strong spatial correlations, thereby generating geographical topological partitions.

[0124] Spatial relationship modeling includes spatial element collection, spatial relationship model construction, spatial association analysis, preliminary topological unit division, unit rationality verification and adjustment, and geographic topological partition generation.

[0125] (3) Adaptive fusion processing of business attribute clustering and geographic topology partitioning is performed to generate a logical segment set.

[0126] For example, an adaptive fusion rule for business attribute clustering and geographic topology partitioning is formulated, clarifying the weight allocation and matching criteria for business homogeneity and geographic relevance. Node groups in the business attribute clustering are compared one by one with units in the geographic topology partitioning to identify the matching parts where nodes belong to the same category and the differences where they belong to conflicting categories.

[0127] For the discrepancies, the priority of business attribute consistency and geospatial relevance is analyzed by combining fusion rules to reconcile conflicting node affiliations and determine the final node grouping. The matching portion and the reconciled discrepancies are then integrated to form a node cluster that combines business homogeneity and geographical relevance, generating a logical segment set.

[0128] The adaptive fusion processing includes fusion rule formulation, clustering and partition comparison, identification of differences, coordination of node affiliation conflicts, integration of node clusters, and generation of logical segment sets.

[0129] In one embodiment, coordination optimization processing is performed based on conflict identification signals to generate coordinated policy instructions, including:

[0130] (1) Perform multi-agent negotiation modeling on the conflict identification signal to generate a conflict resolution strategy space.

[0131] For example, the core information included in the conflict identification signal, such as conflict type, involved logical segments, conflict intensity level, and scope of impact, is analyzed to clarify the goals and constraints of multi-agent negotiation modeling. Based on the number of logical segments involved in the conflict and their business attributes, agents with corresponding functions are configured, with each agent corresponding to the interests of one or a group of logical segments.

[0132] Establish multi-agent negotiation rules, including information exchange mechanisms, norms for expressing interests, conflict compromise thresholds, and the number of negotiation iterations. Through information exchange and interest game among agents, simulate the effects of strategies under different conflict resolution directions, generating a conflict resolution strategy space with multiple solutions, including resource reallocation, execution timing adjustment, process path optimization, and dynamic priority adjustment.

[0133] The multi-agent negotiation modeling process includes parsing the core information of conflict identification signals, clarifying negotiation goals and constraints, configuring agents, formulating negotiation rules, simulating negotiation, and generating conflict resolution strategy space.

[0134] (2) Based on the preset Pareto optimization criterion, the conflict resolution strategy space is subjected to non-dominated solution screening to generate a candidate strategy set.

[0135] For example, the evaluation dimensions corresponding to the predefined Pareto optimization criteria are clearly defined, including core aspects such as resource utilization efficiency, business execution effectiveness, conflict resolution thoroughness, and system operational stability. For each strategy in the conflict resolution strategy space, the effectiveness is quantitatively evaluated from each evaluation dimension, generating multi-dimensional evaluation results for each strategy.

[0136] Based on the Pareto optimization criterion, a comparative analysis of the multi-dimensional evaluation results of all strategies is performed to identify non-dominated solutions where no other strategy outperforms itself in all evaluation dimensions. The validity of the non-dominated solutions is then verified, eliminating solutions that do not meet system constraints or have low execution feasibility. The remaining valid non-dominated solutions are then integrated to generate a candidate strategy set.

[0137] The non-dominated solution screening process includes defining the evaluation dimensions of the Pareto optimization criterion, evaluating the multi-dimensional effects of strategies, identifying non-dominated solutions, verifying the validity of solutions, and generating a set of candidate strategies.

[0138] (3) Perform dynamic priority scheduling on the candidate policy set to generate coordinated policy instructions.

[0139] For example, the evaluation dimensions for dynamic priority scheduling of candidate strategies are determined, including business priority weight, conflict resolution efficiency, resource consumption cost, and the degree of positive impact on the entire supply chain operation. Real-time operational status information of the supply chain network is collected, including real-time resource inventory, order execution progress, and node load, as a dynamic reference for priority scheduling.

[0140] By combining evaluation dimensions and real-time operational status, the priority of each strategy in the candidate strategy set is calculated and ranked to determine the execution priority of each strategy. The strategy with the highest priority is selected, and its operation steps, execution entity, resource configuration standards, and time requirements are refined and transformed into standardized, implementable operation instructions to generate coordinated strategy instructions.

[0141] The dynamic priority scheduling process includes determining the scheduling evaluation dimensions, collecting real-time operating status data, calculating and sorting policy priorities, selecting the optimal policy, refining policy operations, and generating coordinated policy instructions.

[0142] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0143] In one embodiment, such as Figure 3 As shown, this application also provides a digital twin-based real-time supply chain monitoring device 300, which includes:

[0144] The segmented modeling module 301 is used to perform business attribute and geographic topology analysis on the physical supply chain network to obtain a set of logical segments; and to perform digital twin modeling on the set of logical segments to generate a digital twin sub-model.

[0145] The segmented optimization module 302 is used to collect segmented sensor data streams through IoT terminals deployed in each logical segment; perform real-time optimization processing on the segmented sensor data streams; and generate local optimization instruction sets.

[0146] The cross-segment coordination module 303 is used to perform cross-segment conflict detection processing on the local optimization instruction set and generate a conflict identification signal; perform coordination optimization processing based on the conflict identification signal to generate coordinated strategy instructions; and perform execution processing on the coordinated strategy instructions to drive the physical execution terminal.

[0147] The global monitoring module 304 is used to aggregate the segmented indicator vectors output by the digital twin model to generate a full-link operation status diagram; and to analyze and process the full-link operation status diagram based on a preset performance deviation threshold to generate dynamic reorganization instructions.

[0148] Specifically, the segmented modeling module 301 extracts multi-dimensional business features from the nodes of the physical supply chain network, and generates business attribute clusters through standardization and clustering. Simultaneously, it collects spatial elements of the nodes, constructs a spatial relationship model, and divides the geographic topology into partitions. By comparing the two types of results through rule fusion, it coordinates node affiliation to generate a logical segment set. Based on this, it clarifies the physical composition and operational rules for each logical segment, constructs a digital mapping framework, integrates data to establish real-time mapping relationships, completes the simulation capability building, and generates a digital twin sub-model.

[0149] The segmented optimization module 302 deploys adapted IoT terminals for each logical segment, captures information such as physical entity state parameters and operating indicators, and transmits and integrates them to generate a structured segmented sensor data stream. The segmented sensor data stream is cleaned, denoised, and standardized. Combined with business rules and performance standards, it uses digital twin model simulation analysis and optimization deduction to clarify execution requirements such as resource allocation and generate a local optimization instruction set.

[0150] The cross-segment coordination module 303 parses the local optimization instruction set to extract segmented resource demand features, traverses the cross-segment dependency graph to identify conflict node pairs and quantifies their intensity, generating conflict identification signals. Based on the conflict identification signals, multi-agent negotiation modeling is performed to generate a conflict resolution strategy space. Non-dominated solutions are selected according to the Pareto criterion to generate a candidate strategy set. The optimal strategy is selected based on real-time state ranking, generating coordinated strategy instructions. The execution path is defined, and the instructions are issued to physical execution terminals to drive their coordinated operation.

[0151] The global monitoring module 304 performs spatiotemporal alignment of segmented indicator vectors to generate a time-synchronized indicator matrix. It then combines this matrix with topological relationships for weighted fusion to generate a node association strength map, which is subsequently rendered using multimodal methods to create a full-link operational status map. Based on a preset performance deviation threshold, it locates key performance deviation areas within the full-link operational status map. Through topological dependency analysis, it generates segmented impact assessment results and, combined with the threshold, makes adaptive reorganization decisions, generating dynamic reorganization instructions.

[0152] The global monitoring module 304 is also used for:

[0153] Perform multi-dimensional bottleneck localization processing on the end-to-end operational status map to identify key performance deviation areas;

[0154] Topological dependency analysis is performed on key performance deviation regions to generate segmented impact assessment results.

[0155] Adaptive reorganization decision processing is performed based on segmented impact assessment and preset performance deviation thresholds to generate dynamic reorganization instructions.

[0156] The global monitoring module 304 is also used for:

[0157] Multi-level dependency path extraction is performed on key performance deviation areas to generate topological propagation chains;

[0158] The following formula is used to quantify the cascading effects of the topological propagation chain and generate initial influence weights:

[0159]

[0160] in, Indicates the initial influence weight. Indicates the source deviation intensity. Indicates the first The criticality factor of each node This represents the industry experience attenuation coefficient. Indicates the distance from the source node to the... Topological distance between nodes, Indicates the criticality of the node's business. Indicates the node buffer capacity factor. Indicates the total number of nodes in the propagation path;

[0161] The initial influence weights are dynamically adjusted based on the real-time system status to generate the segmented influence evaluation results.

[0162] The global monitoring module 304 is also used for:

[0163] The segmented indicator vectors are aligned in time and space to generate a time-synchronized indicator matrix.

[0164] The time synchronization index matrix is ​​subjected to topological weighted fusion processing to generate a node association strength map;

[0165] Multimodal situation rendering is performed based on node association strength maps to generate a full-link operational situation map.

[0166] The cross-segment coordination module 303 is also used for:

[0167] Perform instruction semantic parsing on the local optimization instruction set to extract segmented resource requirement features;

[0168] Based on the segmented resource demand characteristics, cross-segment dependency graph traversal processing is performed to identify conflict node pairs;

[0169] The conflict intensity of the conflicting node pairs is quantized to generate a conflict identification signal.

[0170] The segmented modeling module 301 is also used for:

[0171] Multi-dimensional business feature extraction is performed on the nodes of the physical supply chain network to generate business attribute clusters;

[0172] Spatial relationship modeling is performed on the nodes of the physical supply chain network to generate geographical topological partitions;

[0173] Adaptive fusion processing is performed on business attribute clustering and geographic topology partitioning to generate a logical segment set.

[0174] The cross-segment coordination module 303 is also used for:

[0175] Multi-agent negotiation modeling is performed on the conflict identification signal to generate a conflict resolution strategy space;

[0176] Based on the preset Pareto optimization criterion, the conflict resolution strategy space is subjected to non-dominated solution screening to generate a candidate strategy set.

[0177] The candidate policy set is dynamically prioritized and scheduled to generate coordinated policy instructions.

[0178] In one embodiment, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0179] In one embodiment, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0180] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0181] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for real-time monitoring of the entire supply chain based on digital twins, characterized in that, The method includes: The physical supply chain network is analyzed for business attributes and geographical topology to obtain a set of logical segments; the set of logical segments is then modeled using digital twins to generate digital twin sub-models. The segmented sensor data streams are collected by IoT terminals deployed in each logical segment; the segmented sensor data streams are optimized in real time to generate local optimization instruction sets; The local optimization instruction set is subjected to cross-segment conflict detection processing to generate a conflict identification signal; based on the conflict identification signal, coordination optimization processing is performed to generate a coordinated strategy instruction; the coordinated strategy instruction is executed to drive the physical execution terminal. The segmented indicator vectors output by the digital twin model are aggregated to generate a full-link operation status map; the full-link operation status map is analyzed and processed based on a preset performance deviation threshold to generate dynamic reorganization instructions.

2. The method for real-time monitoring of the entire supply chain based on digital twins according to claim 1, characterized in that, The analysis and processing of the full-link operation status diagram based on a preset performance deviation threshold to generate dynamic reorganization instructions includes: The entire link operation status map is subjected to multi-dimensional bottleneck location processing to identify key performance deviation areas; Based on the key performance deviation regions, topological dependency analysis is performed to generate segmented impact assessment results. Based on the segmented impact assessment and the preset performance deviation threshold, an adaptive reorganization decision-making process is performed to generate the dynamic reorganization instruction.

3. The method for real-time monitoring of the entire supply chain based on digital twins according to claim 2, characterized in that, The topology dependency analysis based on the key performance deviation region generates segmented impact assessment results, including: Multi-order dependency path extraction is performed on the key performance deviation region to generate a topology propagation chain; The cascading effect of the topological propagation chain is quantified using the following formula to generate initial influence weights: in, Indicates the initial influence weight. Indicates the source deviation intensity. Indicates the first The criticality factor of each node This represents the industry experience attenuation coefficient. Indicates the distance from the source node to the... Topological distance between nodes, Indicates the criticality of the node's business. Indicates the node buffer capacity factor. Indicates the total number of nodes in the propagation path; The initial influence weights are dynamically adjusted based on the real-time system status to generate the segmented influence evaluation results.

4. The method for real-time monitoring of the entire supply chain based on digital twins according to claim 1, characterized in that, The aggregation of the segmented indicator vectors output by the digital twin model to generate a full-link operational status map includes: The segmented index vectors are aligned in time and space to generate a time-synchronized index matrix. The time synchronization index matrix is ​​subjected to topological weighted fusion processing to generate a node association strength map; Multimodal situation rendering is performed based on the node association strength map to generate the full-link operation situation map.

5. The method for real-time monitoring of the entire supply chain based on digital twins according to claim 1, characterized in that, The step of performing cross-segment conflict detection processing on the local optimization instruction set to generate a conflict identification signal includes: The local optimization instruction set is subjected to instruction semantic parsing to extract segmented resource requirement features; Based on the segmented resource demand characteristics, cross-segment dependency graph traversal processing is performed to identify conflicting node pairs; The conflict intensity of the conflicting node pairs is quantized to generate the conflict identification signal.

6. The method for real-time monitoring of the entire supply chain based on digital twins according to claim 1, characterized in that, The process of performing business attribute and geographic topology analysis on the physical supply chain network yields a logical segment set, including: Multi-dimensional business feature extraction processing is performed on the nodes of the physical supply chain network to generate business attribute clusters; Spatial relationship modeling is performed on the nodes of the physical supply chain network to generate geographical topology partitions; The business attribute clustering and the geographic topology partitioning are adaptively fused to generate the logical segment set.

7. The method for real-time monitoring of the entire supply chain based on digital twins according to claim 1, characterized in that, The coordination and optimization process based on the conflict identification signal to generate a coordinated strategy instruction includes: The conflict identification signal is processed by multi-agent negotiation modeling to generate a conflict resolution strategy space; The conflict resolution strategy space is subjected to non-dominated solution screening based on the preset Pareto optimization criterion to generate a candidate strategy set. The candidate strategy set is dynamically prioritized and scheduled to generate the coordinated strategy instruction.

8. A real-time monitoring device for the entire supply chain based on digital twins, characterized in that: The device includes: The segmented modeling module is used to perform business attribute and geographic topology analysis on the physical supply chain network to obtain a set of logical segments; and to perform digital twin modeling on the set of logical segments to generate a digital twin sub-model. The segmented optimization module is used to collect segmented sensor data streams through IoT terminals deployed in each logical segment; perform real-time optimization processing on the segmented sensor data streams, and generate local optimization instruction sets; The cross-segment coordination module is used to perform cross-segment conflict detection processing on the local optimization instruction set and generate a conflict identification signal; perform coordination optimization processing based on the conflict identification signal to generate a coordinated strategy instruction; and perform execution processing on the coordinated strategy instruction to drive the physical execution terminal. The global monitoring module is used to aggregate the segmented indicator vectors output by the digital twin model to generate a full-link operation status map; and to analyze and process the full-link operation status map based on a preset performance deviation threshold to generate dynamic reorganization instructions.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the real-time monitoring method for the entire supply chain based on digital twins as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the real-time monitoring method for the entire supply chain based on digital twins as described in any one of claims 1 to 7.

Images