A supply chain intelligent decision and optimization method and system based on big data driving

Abstract

The application relates to the technical field of supply management, and discloses a supply chain intelligent decision and optimization method and system based on big data driving, which comprises the following steps: structurally associating heterogeneous data of a supply chain to obtain a full-link associated data set; intercepting key events of the full-link associated data set to obtain a standardized data slice of the supply chain; analyzing the network structure of the supply chain based on the running characteristics of the standardized data slice, obtaining the node association relationship, and performing supply-demand matching degree diagnosis to obtain a bottleneck node of the supply chain; performing machine analysis on the bottleneck node to obtain key constraint conditions of the supply chain; based on the key constraint conditions, performing multi-path simulation deduction on historical same-period data and in-transit material information of the supply chain to obtain a decision plan of the supply chain; comprehensively analyzing the decision plan to obtain an optimization decision of the supply chain; and the application can improve the efficiency of the supply chain intelligent decision and optimization based on big data driving.

Description

Technical Field

[0001] This invention relates to the field of supply management technology, and in particular to a big data-driven intelligent decision-making and optimization method and system for the supply chain. Background Technology

[0002] In current supply chain management processes, data from multiple points is heterogeneously distributed, and there is a lack of effective structured correlation and fusion methods for data at each stage. This makes it impossible to form a complete, interconnected dataset, hindering a comprehensive and refined analysis of the supply chain's operational status. Furthermore, the lack of standardized methods for data slicing and feature extraction makes it difficult to accurately capture key characteristics of supply chain operations, effectively analyze node relationships and supply-demand matching, and result in low efficiency and accuracy in locating bottleneck nodes.

[0003] Current technologies for analyzing supply chain bottlenecks only identify surface-level phenomena, lacking in-depth analysis of their formation mechanisms and failing to clearly define the key constraints on supply chain operation. Decision-making based on a single data dimension struggles to combine historical data and information on goods in transit for multi-path simulations. The generation of decision-making plans lacks scientific rigor and foresight, and subsequent assessments lack systematic performance evaluation and conflict resolution mechanisms, resulting in insufficient rationality and feasibility of supply chain decisions. Therefore, improving the efficiency and accuracy of intelligent supply chain decision-making has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a big data-driven intelligent decision-making and optimization method and system for the supply chain to solve the problems mentioned in the background.

[0005] To achieve the above objectives, this invention provides a big data-driven intelligent decision-making and optimization method for the supply chain, comprising: S1. Integrate the aviation material inventory data, global positioning data, environmental perception data and settlement cycle data of the supply chain into heterogeneous data, and perform structured association on the heterogeneous data to obtain the full-link association dataset of the supply chain. S2. Extract key events from the full-link associated dataset to obtain standardized data slices of the supply chain; S3. Extract the operational characteristics of the standardized data slices, and based on the operational characteristics, perform network structure analysis on the supply chain to obtain the node relationships of the supply chain, and perform supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain. S4. Perform mechanism analysis on the bottleneck node to obtain the key constraints of the supply chain; S5. Based on the key constraints, perform multi-path simulation and deduction on the historical data of the same period of the supply chain and the information of goods in transit to obtain the decision-making plan for the supply chain. S6. Conduct a comprehensive analysis of the decision-making plan to obtain the optimized decision for the supply chain.

[0006] In a preferred embodiment, the process of integrating supply chain data on aircraft material inventory, global positioning data, environmental sensing data, and settlement cycle data into heterogeneous data, and then performing structured correlation on the heterogeneous data to obtain a full-link correlation dataset of the supply chain, includes: Semantic parsing of fields is performed on aviation material inventory data from the procurement side, global positioning data from the transportation side, environmental perception data from the warehousing side, and settlement cycle data from the financial side to obtain the initial structured record of the supply chain. Based on the initial structured records, the task data of the supply chain is cross-source aligned to obtain the task-level intermediate data of the supply chain; The task-level intermediate data is fused over time to obtain the time-series correlation data of the supply chain; Spatial registration is performed on the time-series correlated data to obtain the spatial alignment data of the supply chain; The time-series correlated data and the spatially aligned data are fused and stitched together to obtain the full-link correlated dataset of the supply chain.

[0007] In a preferred embodiment, the step of extracting key events from the end-to-end related dataset to obtain standardized data slices of the supply chain includes: By identifying the operational nodes in the full-link associated dataset, the key event types of the supply chain can be obtained; Based on the key event types, time-series fluctuations are captured in the full-link associated dataset to obtain the event timestamps of the supply chain; Based on the timestamp, the contextual traceability of the full-link associated dataset is extended to obtain dynamic data fragments of the supply chain; Field filtering is performed on the dynamic data fragments to obtain standardized data slices of the supply chain.

[0008] In a preferred embodiment, the step of extracting the operational characteristics of the standardized data slices, performing network structure analysis on the supply chain based on the operational characteristics to obtain the node relationships of the supply chain, and performing supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain includes: The standardized data slices are subjected to time-series analysis to obtain their dynamic operating characteristics. Based on the dynamic operation characteristics, an association topology is constructed with the nodes of the supply chain as entities and the node flow paths of the supply chain as boundaries. The flow density of the edges in the association topology is evaluated to obtain the node association relationship of the supply chain. By comparing the supply and demand scales of the associated nodes in the node association relationship, the supply and demand matching deviation of the associated nodes is obtained; Based on the supply and demand mismatch, the associated nodes are traversed and screened to obtain the bottleneck nodes of the supply chain.

[0009] In a preferred embodiment, based on the dynamic operating characteristics, constructing an association topology with the nodes of the supply chain as entities and the node flow paths of the supply chain as boundaries, and evaluating the flow tightness of the edges in the association topology to obtain the node association relationships of the supply chain includes: Based on the dynamic operation characteristics, business nodes and material flow paths in the supply chain are identified, and the relationship between the business nodes and the material flow paths is deduced to obtain the node and path relationship of the supply chain. Based on the relationship between the nodes and paths, an association topology is constructed with the business nodes as entities and the material flow paths as boundaries; The frequency of material interactions is statistically analyzed on the associated topology to obtain the interaction frequency feature value of the material flow path; The timeliness of the material flow path is compared by performing a flow timeliness comparison on the associated topology to obtain the timeliness stability characteristic value of the material flow path; The interaction frequency feature value and the timeliness stability feature value are coupled and evaluated to obtain the node correlation tightness of the supply chain; The node associations are sorted and categorized based on their degree of association to obtain the node association relationships in the supply chain.

[0010] In a preferred embodiment, the formula for calculating the node association tightness is as follows: ; in, The node association tightness, The interaction frequency feature value is... The actual circulation time of the aforementioned time-stability characteristic value. The historical average turnover time of the aforementioned time-stability characteristic value. The preset weighting coefficients, The preset weighting coefficients, It is an exponential function.

[0011] In a preferred embodiment, the step of performing mechanistic analysis on the bottleneck node to obtain the key constraints of the supply chain includes: By tracing the upstream and downstream supply and demand relationships of the bottleneck node, the supply and demand imbalance boundary of the bottleneck node can be obtained. The operational capacity of the bottleneck node is analyzed to obtain the capacity carrying threshold of the bottleneck node. The bottleneck node is analyzed for external environment correlation to obtain the environmental constraints of the bottleneck node. By attributing and aggregating the supply and demand imbalance boundary, the capacity carrying capacity threshold, and the environmental constraints, the key constraints of the supply chain are obtained.

[0012] In a preferred embodiment, the step of performing multi-path simulation and extrapolation based on the key constraints and historical contemporaneous data and in-transit goods information of the supply chain to obtain a decision-making plan for the supply chain includes: Retrieve historical data from the same period in the supply chain to obtain the historical benchmark for the supply chain; The status of in-transit goods in the supply chain is analyzed to obtain the status of in-transit goods in the supply chain. Based on the aforementioned key constraints, a perturbation injection is performed on the historical reference benchmark to obtain the first set of evolution path branches of the supply chain; Based on the aforementioned key constraints, the status of the goods in transit is prospectively extended to obtain the second set of evolution path branches of the supply chain. The first set of evolution path branches and the second set of evolution path branches are merged and deduplicated to obtain the decision plan for the supply chain.

[0013] In a preferred embodiment, the step of comprehensively evaluating the decision-making plans to obtain the optimized decision for the supply chain includes: The effectiveness indicators of the decision-making plan are extracted to obtain the multidimensional characteristics of the path effectiveness of the supply chain; By performing statistical analysis on the interrelationships of the efficiency dimensions of the multidimensional features, the efficiency correlation of the supply chain can be obtained. Based on the aforementioned performance correlation, scenario stress tests are conducted on the decision-making plan to obtain the scenario tolerance of the decision-making plan. Based on the scenario tolerance, the decision-making plan is analyzed at the operational micro level to obtain the resource conflict list of the decision-making plan; Based on the resource conflict list, conflict resolution is performed on the decision-making plan to obtain a feasible plan. The execution complexity of the feasible contingency plan is evaluated to obtain the execution smoothness of the decision-making contingency plan; The execution smoothness and the scenario tolerance are fused and encapsulated to obtain the optimized decision for the supply chain.

[0014] To address the aforementioned problems, this invention also provides a big data-driven intelligent supply chain decision-making and optimization system, the system comprising: The end-to-end data association module is used to integrate aviation material inventory data, global positioning data, environmental perception data and settlement cycle data of the supply chain into heterogeneous data, and to perform structured association on the heterogeneous data to obtain the end-to-end association dataset of the supply chain. The standardized data slicing module is used to extract key events from the full-link associated dataset to obtain standardized data slices of the supply chain. The bottleneck node diagnosis module is used to extract the operational characteristics of the standardized data slices, and based on the operational characteristics, perform network structure analysis on the supply chain to obtain the node relationships of the supply chain, and perform supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain. The mechanism analysis module is used to perform mechanism analysis on the bottleneck node to obtain the key constraints of the supply chain. The decision-making plan generation module is used to perform multi-path simulation and deduction based on the key constraints, historical data of the same period of the supply chain and information on goods in transit, to obtain the decision-making plan for the supply chain. The comprehensive analysis module is used to comprehensively analyze the decision-making plan to obtain the optimized decision for the supply chain.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates heterogeneous data from multiple ends of the supply chain and completes structured association. By combining key event interception and operational feature extraction, it achieves precise analysis of the supply chain network structure, efficiently locates bottleneck nodes and analyzes core constraints, and conducts multi-path simulation and deduction based on historical data and information on goods in transit. This makes the generation of supply chain decision-making plans more targeted and scientific, significantly improving the accuracy and rationality of supply chain decisions. It realizes intelligent processing of the entire chain from data integration to node diagnosis to plan generation, effectively enhancing the efficiency and depth of data utilization and analysis in the supply chain decision-making process.

[0016] 2. This invention establishes a supply chain intelligent decision-making and optimization system. Through the collaborative operation of various functional modules, it extracts the effectiveness of decision-making plans, performs stress testing, and resolves conflicts. It can screen out optimization decisions with excellent execution smoothness and scenario tolerance, significantly improving the feasibility and adaptability of supply chain optimization decisions. At the same time, by using standardized data processing procedures and quantitative node correlation analysis methods, it makes supply chain operation optimization more operable, effectively improving the overall operational efficiency and resource allocation efficiency of the supply chain, and realizing the intelligent and systematic upgrade of supply chain decision-making and optimization. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a big data-driven intelligent decision-making and optimization method for the supply chain is provided in one embodiment of the present invention. Figure 2 A functional block diagram of a big data-driven intelligent supply chain decision-making and optimization system provided in an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a big data-driven intelligent supply chain decision-making and optimization method. The executing entity of this big data-driven intelligent supply chain decision-making and optimization method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the big data-driven intelligent supply chain decision-making and optimization method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a big data-driven intelligent supply chain decision-making and optimization method according to an embodiment of the present invention. In this embodiment, the big data-driven intelligent supply chain decision-making and optimization method includes: S1. Integrate the aviation material inventory data, global positioning data, environmental perception data and settlement cycle data of the supply chain into heterogeneous data, and perform structured association on the heterogeneous data to obtain the full-link association dataset of the supply chain. In this embodiment of the invention, the process of integrating supply chain data such as aircraft material inventory, global positioning data, environmental sensing data, and settlement cycle data into heterogeneous data, and then performing structured correlation on the heterogeneous data to obtain the full-link correlation dataset of the supply chain, includes: Semantic parsing of fields is performed on aviation material inventory data from the procurement side, global positioning data from the transportation side, environmental perception data from the warehousing side, and settlement cycle data from the financial side to obtain the initial structured record of the supply chain. Based on the initial structured records, the task data of the supply chain is cross-source aligned to obtain the task-level intermediate data of the supply chain; The task-level intermediate data is fused over time to obtain the time-series correlation data of the supply chain; Spatial registration is performed on the time-series correlated data to obtain the spatial alignment data of the supply chain; The time-series correlated data and the spatially aligned data are fused and stitched together to obtain the full-link correlated dataset of the supply chain.

[0021] Data is collected from the procurement end (aviation material inventory), the transportation end (global positioning data), the warehousing end (environmental awareness data), and the financial end (settlement cycle data). Semantic analysis of the fields is conducted, and the field names, meanings, and data types of each field are identified and defined to clarify the business attributes and data references of each field. Invalid and missing fields in each field are marked and processed. The data from each field is then reorganized and entered according to a unified field description standard to form a structured dataset, which serves as the initial structured record of the supply chain.

[0022] The system matches and associates relevant data for the same task in the initial structured records of procurement, transportation, warehousing, and finance. It unifies the format of heterogeneous data for the same task from different data sources, eliminating differences in data format and expression between different data sources. It integrates all the data for the same task after association, matching, and processing into a complete task data record. All the integrated task data records together constitute the task-level intermediate data of the supply chain.

[0023] Extract the time attribute information from each task data record, sort all task data records in chronological order, connect and integrate task data records from different stages of the same business process according to the time axis, supplement the correlation information of each task data record in the time dimension, so that the integrated dataset can fully reflect the flow process of the supply chain business in the time dimension. This integrated dataset is the time-series correlation data of the supply chain.

[0024] Extract all information involving physical location from the time-series correlation data, and use a unified spatial coordinate system to transform and calibrate all location information. Clarify the specific location and relative relationship of each business node and material transportation path in the spatial coordinate system. Bind the spatial information after location calibration with the corresponding time-series correlation data so that each data has a precise spatial location identifier. The bound dataset is the spatial alignment data of the supply chain.

[0025] Based on time-series correlated data as the basic framework, the spatial location information corresponding to the spatial alignment data is fully mapped to each data record of the time-series correlated data. At the same time, the time dimension information of the time-series correlated data and the spatial dimension information of the spatial alignment data are retained, realizing the complete integration of supply chain data in both time and space dimensions. The complete dataset formed after fusion and splicing, which includes time, space and business attributes of each end, is the full-link correlated dataset of the supply chain.

[0026] The beneficial effects are that by completing the parsing, alignment, fusion, and splicing of heterogeneous data from multiple ends in stages, the effective integration and structured association of data from procurement, transportation, warehousing, and finance ends are achieved. This transforms the originally scattered heterogeneous data into a full-link associated dataset with both time and spatial dimensions, fully preserving the temporal flow information and spatial location information of the supply chain business, eliminating information barriers between data from different ends, and providing complete, accurate, and related basic data support for subsequent data analysis, node diagnosis, and decision optimization of the entire supply chain, thus ensuring the comprehensiveness and accuracy of subsequent supply chain-related analysis work.

[0027] S2. Extract key events from the full-link associated dataset to obtain standardized data slices of the supply chain; In this embodiment of the invention, the step of extracting key events from the end-to-end related dataset to obtain standardized data slices of the supply chain includes: By identifying the operational nodes in the full-link associated dataset, the key event types of the supply chain can be obtained; Based on the key event types, time-series fluctuations are captured in the full-link associated dataset to obtain the event timestamps of the supply chain; Based on the timestamp, the contextual traceability of the full-link associated dataset is extended to obtain dynamic data fragments of the supply chain; Field filtering is performed on the dynamic data fragments to obtain standardized data slices of the supply chain.

[0028] By analyzing all supply chain operational links in the full-link associated dataset, clarifying the operational node attributes corresponding to each operational link, and classifying and defining the business events corresponding to each operational node based on the core processes and key control points of supply chain business operations, the event categories that can reflect the key status of supply chain operation are determined, and these categories are the key event types of the supply chain.

[0029] Extract all data information that matches the key event type from the full-link associated dataset, monitor the numerical changes and status anomalies of this data in the time dimension, locate the specific time points when the data changes and status anomalies occur, record and label the specific time points according to a unified time format, and the recorded and labeled time information is the event timestamp of the supply chain.

[0030] Using the timestamp of each event as the core time point, we trace back the supply chain-related data within a preset time period before that time point and extend the supply chain-related data within a preset time period after that time point. We integrate the key data of the core time point with the related data obtained from tracing and extension to form a continuous data set that includes the entire process of the event. This data set is the dynamic data segment of the supply chain.

[0031] By sorting out all data fields contained in the dynamic data fragments, and based on the actual needs of supply chain critical event analysis, the core data fields that need to be retained are determined. Redundant fields that are irrelevant to critical event analysis are removed from the dynamic data fragments. The retained core data fields are arranged in a uniform field order. The standardized dataset formed after the arrangement is the standardized data slice of the supply chain.

[0032] The beneficial effects are that, through step-by-step operations such as job node identification, time-series fluctuation capture, context tracing and extension, and field filtering, accurate key event extraction of the entire chain-related dataset is achieved. The resulting standardized data slices eliminate redundant data while retaining complete information on key events, allowing supply chain data analysis to focus on core business events. This provides a standardized, accurate, and targeted data source for subsequent extraction of operational features and analysis of network structure, effectively improving the efficiency and accuracy of subsequent supply chain data analysis. At the same time, the standardized processing method ensures that data slices of different key events have a unified format, facilitating subsequent comparative analysis and feature extraction.

[0033] S3. Extract the operational characteristics of the standardized data slices, and based on the operational characteristics, perform network structure analysis on the supply chain to obtain the node relationships of the supply chain, and perform supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain. In this embodiment of the invention, the step of extracting the operational characteristics of the standardized data slices, performing network structure analysis on the supply chain based on the operational characteristics to obtain the node relationships of the supply chain, and performing supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain includes: The standardized data slices are subjected to time-series analysis to obtain their dynamic operating characteristics. Based on the dynamic operation characteristics, an association topology is constructed with the nodes of the supply chain as entities and the node flow paths of the supply chain as boundaries. The flow density of the edges in the association topology is evaluated to obtain the node association relationship of the supply chain. By comparing the supply and demand scales of the associated nodes in the node association relationship, the supply and demand matching deviation of the associated nodes is obtained; Based on the supply and demand mismatch, the associated nodes are traversed and screened to obtain the bottleneck nodes of the supply chain.

[0034] Based on the dynamic operating characteristics, a relational topology is constructed with the nodes of the supply chain as entities and the node flow paths of the supply chain as boundaries. The flow density of the edges in the relational topology is evaluated to obtain the node association relationships of the supply chain, including: Based on the dynamic operation characteristics, business nodes and material flow paths in the supply chain are identified, and the relationship between the business nodes and the material flow paths is deduced to obtain the node and path relationship of the supply chain. Based on the relationship between the nodes and paths, an association topology is constructed with the business nodes as entities and the material flow paths as boundaries; The frequency of material interactions is statistically analyzed on the associated topology to obtain the interaction frequency feature value of the material flow path; The timeliness of the material flow path is compared by performing a flow timeliness comparison on the associated topology to obtain the timeliness stability characteristic value of the material flow path; The interaction frequency feature value and the timeliness stability feature value are coupled and evaluated to obtain the node correlation tightness of the supply chain; The node associations are sorted and categorized based on their degree of association to obtain the node association relationships in the supply chain.

[0035] The formula for calculating the node association density is as follows: ; in, The node association tightness, The interaction frequency feature value is... The actual circulation time of the aforementioned time-stability characteristic value. The historical average turnover time of the aforementioned time-stability characteristic value. The preset weighting coefficients, The preset weighting coefficients, It is an exponential function.

[0036] By sorting out the time series information of all data in the standardized data slices, the operation process of the supply chain business is broken down according to the time progression. Information on the dynamic changes of the supply chain, such as the status of material flow, node operation status, and path passage status at different time stages, is extracted. All dynamic change information is integrated to form a feature set that can reflect the real-time operation status of the supply chain. This feature set is the dynamic operation feature of the standardized data slice.

[0037] The system extracts business node information that identifies each operational link in the supply chain from dynamic operational characteristics. At the same time, it extracts information on the material transportation, allocation and other circulation paths that connect each business node. It sorts out all material circulation paths corresponding to each business node and clarifies the upstream and downstream business nodes corresponding to each material circulation path. This sorts out a system of correspondence between business nodes and material circulation paths, which is the node and path relationship of the supply chain.

[0038] Each business node is treated as an independent entity node in the associated topology, and each material flow path is treated as a boundary edge connecting the corresponding entity node. According to the correspondence between the node and the path, all entity nodes are connected to the boundary edges to form a topology that can intuitively reflect the relationship between the business nodes and the flow paths in the supply chain. This topology is the associated topology of the supply chain.

[0039] Using a preset statistical period as the time range, the number of material interactions is counted for each material flow path corresponding to each boundary edge in the associated topology. The total number of material interactions for each material flow path within the statistical period is recorded, and this total number is used as a quantitative indicator and assigned to the corresponding material flow path. This quantitative indicator is the interaction frequency characteristic value of the material flow path.

[0040] Extract the actual material circulation time data of each boundary edge in the associated topology corresponding to the material circulation path, and at the same time retrieve the historical material circulation time data of the path and calculate the average value. Compare and analyze the actual material circulation time with the historical average circulation time to form a quantitative indicator that can reflect the stability of the circulation time of the path. This quantitative indicator is the timeliness stability characteristic value of the material circulation path.

[0041] According to the preset weight allocation rules, weight coefficients are assigned to the interaction frequency feature value and the timeliness stability feature value respectively. After multiplying each feature value by the corresponding weight coefficient, the sum is calculated to obtain a quantitative value that can comprehensively reflect the degree of correlation of the material flow path. This value is the node correlation tightness of the supply chain.

[0042] All nodes corresponding to the material flow path are arranged in descending order of their correlation strength. Based on the preset numerical range, the correlation strength of the sorted nodes is divided into different levels. Each level corresponds to a different degree of correlation between business nodes. Combining the level division results with the corresponding business nodes and material flow path information, a complete supply chain node correlation system is formed. This system is the node correlation relationship of the supply chain.

[0043] Based on dynamic operational characteristics, business nodes and material flow paths in the supply chain are identified. Relationships between these nodes and paths are deduced, and a topology is constructed with business nodes as entities and material flow paths as boundaries. Material interaction frequency is statistically analyzed on this topology to directly obtain interaction frequency feature values. The flow timeliness of the topology is compared to obtain the actual flow timeliness and the historical average flow timeliness of the timeliness stability feature values. The preset weighting coefficients are fixed values ​​pre-set manually based on the actual operational needs and business characteristics of the supply chain. The exponential function is a general mathematical exponential function, and its calculation rule is based on natural constants. The specified exponentiation operation.

[0044] By subtracting the natural constant from the interaction frequency characteristic value. The negative interaction frequency characteristic value is raised to a power, then multiplied by a preset first weighting coefficient. The absolute value of the actual circulation time is subtracted from the historical average circulation time. The result is divided by the historical average circulation time, and the result is then rounded to the natural constant. The operation of raising the negative value to a power, multiplying the result with a preset second weight coefficient, and finally adding the results of the two multiplications to obtain the node association tightness. This calculation process can couple and evaluate the characteristics of two dimensions of material flow path: interaction frequency and timeliness stability, and quantitatively reflect the degree of association between two business nodes in the supply chain through the corresponding material flow path, so as to achieve accurate quantitative judgment of the relationship between supply chain nodes.

[0045] When the interaction frequency characteristic value of the material flow path increases, the natural constant... The result of raising the negative interaction frequency characteristic value to a power will decrease, the value of 1 minus the result will increase, and the result after multiplying by the preset first weight coefficient will also increase. Ultimately, the node correlation density will show an increasing trend. When the absolute value of the difference between the actual circulation time and the historical average circulation time decreases, the ratio of this absolute value to the historical average circulation time will decrease, and the natural constant... The result of raising the negative power of the ratio will increase, and the result after multiplying it with the preset second weight coefficient will also increase. Ultimately, the node association tightness will show an increasing trend. Conversely, when the interaction frequency feature value decreases, or when the absolute value of the difference between the actual circulation time and the historical average circulation time increases, the node association tightness will show a decreasing trend. The two preset weight coefficients are fixed values, and their values ​​will not change with the feature value. They are only used as the basis for adjusting the ratio of the two-dimensional calculation results to ensure the uniformity of the calculation results and the adaptability to actual business.

[0046] Based on the supply and demand matching deviation, the associated nodes are traversed and screened. A reasonable numerical range for supply and demand matching is set. The supply and demand matching deviation of each group of associated nodes is compared with the reasonable numerical range. Associated nodes whose deviation exceeds the reasonable numerical range are marked. All associated nodes are traversed in turn to complete the comparison and marking. All marked associated nodes are integrated. The integrated node set is the bottleneck node of the supply chain.

[0047] The beneficial effects include the accurate extraction of dynamic operation characteristics of the supply chain from standardized data slices through time-series analysis, the construction of the associated topology and the multi-dimensional assessment of the tightness of flow based on these characteristics, the quantification and classification of node relationships, and the determination of supply-demand matching deviation through supply-demand scale comparison. Based on this, the accurate screening of bottleneck nodes is completed. The entire process is progressive, deeply integrating supply chain network structure analysis with supply-demand matching diagnosis. This not only achieves the scientific quantification of node relationships but also accurately locates bottleneck nodes in the supply chain that are imbalanced between supply and demand. This provides precise node guidance for subsequent analysis of the mechanism of supply chain bottlenecks and decision optimization, significantly improving the accuracy and pertinence of supply chain network analysis and ensuring that subsequent optimization decisions can focus on core problem nodes.

[0048] S4. Perform mechanism analysis on the bottleneck node to obtain the key constraints of the supply chain; In this embodiment of the invention, the step of performing mechanistic analysis on the bottleneck node to obtain the key constraints of the supply chain includes: By tracing the upstream and downstream supply and demand relationships of the bottleneck node, the supply and demand imbalance boundary of the bottleneck node can be obtained. The operational capacity of the bottleneck node is analyzed to obtain the capacity carrying threshold of the bottleneck node. The bottleneck node is analyzed for external environment correlation to obtain the environmental constraints of the bottleneck node. By attributing and aggregating the supply and demand imbalance boundary, the capacity carrying capacity threshold, and the environmental constraints, the key constraints of the supply chain are obtained.

[0049] By identifying the upstream supply nodes and downstream demand nodes corresponding to the bottleneck node, extracting material supply data, material demand data, and material flow data between each related node and the bottleneck node, clarifying the position and role of the bottleneck node in the supply chain, defining the upper limit of material supply and the lower limit of material demand for the bottleneck node, and defining the specific numerical range and boundary conditions for the imbalance between supply and demand at the bottleneck node. These numerical ranges and boundary conditions are the supply and demand imbalance boundaries of the bottleneck node.

[0050] By analyzing all information related to operational capacity, such as the configuration of equipment, personnel, site size, and operational procedures, at the bottleneck node, and combining this with actual operational indicators such as data processing efficiency, material handling speed, and operational error tolerance, the maximum scale of materials that the bottleneck node can process and the highest frequency of operations that it can complete per unit time are determined through a comprehensive calculation of hardware configuration capabilities and actual operational capabilities. These maximum scale and highest frequency are the capacity carrying thresholds of the bottleneck node.

[0051] We analyze various external environmental factors that affect the normal operation of bottleneck nodes, including natural environment, policy environment, market environment, transportation environment, etc. We extract data on the scope, intensity, and duration of influence of each external environmental factor, and clarify the specific constraints and degree of each external environmental factor on the operational efficiency, material flow, and supply and demand matching of bottleneck nodes. These specific constraints and degree of influence are the environmental constraints of bottleneck nodes.

[0052] By attributing and aggregating supply and demand imbalance boundaries, capacity carrying capacity thresholds, and environmental constraints, we can identify the causes of bottleneck problems corresponding to each of these factors. We can also analyze the internal connections and mutual influences among these three factors. By integrating and summarizing the scattered constraint information from various dimensions, we can form a complete information set that comprehensively reflects the causes and operational constraints of bottleneck nodes. This complete information set is the key constraint condition of the supply chain.

[0053] The beneficial effects are that by tracing the upstream and downstream supply and demand relationships, analyzing operational capabilities, and interpreting the external environment of bottleneck nodes, the core constraints of bottleneck nodes can be accurately identified from the three dimensions of supply and demand, operation, and environment. Then, by attribution aggregation, the constraint information of each dimension is integrated, and the resulting key constraints can comprehensively and accurately reflect the core limitations of supply chain operation. This provides accurate and comprehensive constraint basis for subsequent multi-path simulation and deduction based on key constraints, ensuring that the subsequent decision-making plans are closely aligned with the actual operational limitations of the supply chain and guaranteeing the scientificity and feasibility of the decision-making plans.

[0054] S5. Based on the key constraints, perform multi-path simulation and deduction on the historical data of the same period of the supply chain and the information of goods in transit to obtain the decision-making plan for the supply chain. In this embodiment of the invention, the step of performing multi-path simulation and deduction based on the key constraints and historical data of the supply chain and information on goods in transit to obtain a decision-making plan for the supply chain includes: Retrieve historical data from the same period in the supply chain to obtain the historical benchmark for the supply chain; The status of in-transit goods in the supply chain is analyzed to obtain the status of in-transit goods in the supply chain. Based on the aforementioned key constraints, a perturbation injection is performed on the historical reference benchmark to obtain the first set of evolution path branches of the supply chain; Based on the aforementioned key constraints, the status of the goods in transit is prospectively extended to obtain the second set of evolution path branches of the supply chain. The first set of evolution path branches and the second set of evolution path branches are merged and deduplicated to obtain the decision plan for the supply chain.

[0055] Based on the time cycle of supply chain operations, historical operational data matching the current analysis period is extracted, covering all dimensions of supply chain data such as supply and demand data, material flow data, node operation data, and environmental impact data within this period. The extracted historical data undergoes integrity verification and standardization processing, invalid data is removed, and missing key information is supplemented. After processing, a complete dataset that reflects the historical operating status of the supply chain during the same period is formed. This dataset serves as the historical reference benchmark for the supply chain.

[0056] The system comprehensively collects basic information on goods in transit, including their type, quantity, transportation route, current location, transportation timeliness, and loading / unloading nodes. It also combines global positioning data and environmental perception data from the transportation end to analyze the real-time transportation status, potential transportation risks, and estimated arrival time of goods in transit. By integrating all collected information and analysis results, a comprehensive information set that reflects the current and imminent future status of goods in transit is formed. This information set is the status of goods in transit in the supply chain.

[0057] Perturbation injection is performed on historical reference benchmarks. The supply and demand imbalance boundary, capacity carrying capacity threshold, and environmental constraints contained in the key constraints are used as perturbation factors. Perturbation factors are injected into the historical reference benchmarks one by one according to different degrees of influence and modes of action. The operation and evolution of the supply chain under the individual action of each perturbation factor and the combined action of multiple factors are simulated. The operation path and development results of the supply chain under each perturbation scenario are recorded. All recorded paths and results together constitute the first set of evolution path branches of the supply chain.

[0058] Based on the status of goods in transit and combined with various operational restrictions set by key constraints, the system extrapolates the operational status of goods in transit in subsequent transportation, warehousing, and handover stages according to the time sequence and spatial path of goods flow. It simulates the flow results of goods in transit and the overall operational changes of the supply chain under different response measures, and sorts out a variety of different paths for the flow of goods and the operation and development of the supply chain. All the sorted paths together constitute the second set of evolutionary path branches of the supply chain.

[0059] The path information of all paths in the two sets of evolution path branches is compared one by one to identify duplicate paths with completely identical path content, operation results and implementation methods. All duplicate paths are eliminated and only one valid path is retained. The remaining valid paths in the two sets of path branches are integrated and summarized to form a path set containing a variety of different supply chain operation solutions. This path set is the decision-making plan for the supply chain.

[0060] The beneficial effects are as follows: by retrieving historical data from the same period to construct a historical reference benchmark, and analyzing information on goods in transit to form a status quo of goods in transit, a two-dimensional basis for multi-path simulation and extrapolation that fits the actual operation of the supply chain is provided. By combining key constraints to carry out disturbance injection and forward-looking extension, two sets of evolution path branches are generated respectively, so that the simulated paths are both in line with historical experience and the current actual status of goods. Then, by merging and deduplicating to form decision-making plans, the diversity and comprehensiveness of decision-making plans are ensured, while redundant paths are eliminated. This provides accurate, rich and non-redundant decision-making references for subsequent comprehensive analysis, making the generation of supply chain decision-making plans more scientific, targeted and practical.

[0061] S6. Conduct a comprehensive analysis of the decision-making plan to obtain the optimized decision for the supply chain.

[0062] In this embodiment of the invention, the step of comprehensively evaluating the decision-making plan to obtain the optimized decision for the supply chain includes: The effectiveness indicators of the decision-making plan are extracted to obtain the multidimensional characteristics of the path effectiveness of the supply chain; By performing statistical analysis on the interrelationships of the efficiency dimensions of the multidimensional features, the efficiency correlation of the supply chain can be obtained. Based on the aforementioned performance correlation, scenario stress tests are conducted on the decision-making plan to obtain the scenario tolerance of the decision-making plan. Based on the scenario tolerance, the decision-making plan is analyzed at the operational micro level to obtain the resource conflict list of the decision-making plan; Based on the resource conflict list, conflict resolution is performed on the decision-making plan to obtain a feasible plan. The execution complexity of the feasible contingency plan is evaluated to obtain the execution smoothness of the decision-making contingency plan; The execution smoothness and the scenario tolerance are fused and encapsulated to obtain the optimized decision for the supply chain.

[0063] By analyzing all business indicators corresponding to the supply chain operation paths in the decision-making plan, core performance indicators including material flow efficiency, node operation efficiency, resource utilization efficiency, supply and demand matching accuracy, and operating cost control are extracted. Each core performance indicator is quantified and standardized, and all quantified performance indicators are integrated to form a multi-dimensional feature set, which is the multi-dimensional feature of supply chain path performance.

[0064] By sorting out the quantitative data of each efficiency dimension in the multidimensional characteristics of path efficiency, statistically analyzing the numerical change relationship between any two efficiency dimensions, calculating the correlation value between each efficiency dimension, clarifying the quantitative relationship of mutual influence and mutual constraint among each efficiency dimension, and integrating the correlation values ​​of all efficiency dimensions to form a complete correlation system, which is the efficiency correlation of the supply chain.

[0065] Based on the correlation between various efficiency dimensions in the efficiency correlation, various stress scenarios that may occur in the supply chain operation process, such as extreme market environments, sudden situations in material flow, and node operation failures, are simulated. Each stress scenario is converted into numerical fluctuations of the corresponding efficiency dimension. The numerical fluctuations are substituted into the decision-making plan and the overall operation status of the plan is deduced. The adaptability and operational stability of each decision-making plan under different stress scenarios are determined. The quantitative results of this adaptability and operational stability are the scenario tolerance of the decision-making plan.

[0066] Decision-making plans that meet the preset standards for scenario tolerance are selected. The specific execution steps of each plan are broken down according to the supply chain operation process. The specific configuration and usage information of various resources such as manpower, material resources, venues, and equipment required in each step of the execution process are sorted out. Problems such as resource overlap, resource shortage, and resource usage conflict between different execution steps and different resource configurations are identified. All identified problems are integrated and recorded to form a list, which is the resource conflict list of the decision-making plan.

[0067] For the various resource conflict issues recorded in the resource conflict list, and in combination with the actual operating resources and business needs of the supply chain, corresponding conflict resolution strategies such as resource allocation, process adjustment, and node coordination are formulated. The resolution strategies are applied to the corresponding decision-making plans one by one, and the execution steps and resource allocation of the plans are optimized and adjusted. Plans that still cannot resolve conflicts after adjustment are eliminated, and all plans that effectively resolve all conflict issues are retained. The retained plans are the feasible plans for decision-making.

[0068] The number of work nodes, process connection links, number of resource allocations, and frequency of cross-terminal collaborations during the execution of each contingency plan are statistically analyzed. Based on the preset complexity assessment rules, each quantitative indicator is weighted and calculated to obtain a quantitative value of execution complexity. This value is then reverse-converted to obtain a quantitative result that reflects the ease of execution of the contingency plan. This result is the smoothness of the decision contingency plan's execution.

[0069] The execution smoothness and scenario tolerance are integrated and encapsulated. Preset weights that conform to the actual business needs of the supply chain are configured for execution smoothness and scenario tolerance respectively. The quantitative values ​​of execution smoothness and scenario tolerance of each feasible plan are multiplied by the corresponding weights respectively. The results of the two multiplications are summed to obtain a comprehensive evaluation value. Feasible plans that meet the preset optimal standard are selected. The plan is then refined in detail based on the actual operation requirements of the supply chain. The refined plan is the optimization decision of the supply chain.

[0070] The beneficial effects are that through a series of step-by-step operations such as performance indicator extraction, performance relationship statistics, and scenario stress testing, a multi-dimensional and comprehensive assessment of decision-making plans has been achieved. This approach analyzes the operational value of the plans from the performance perspective, verifies their adaptability from the stress perspective, resolves resource conflicts and assesses ease of execution from the execution perspective, and finally selects optimized decisions that balance scenario resilience and smooth execution through fusion and encapsulation. This ensures that the final decisions of the supply chain are not only resilient to various emergencies but also convenient and feasible to implement, effectively improving the scientific nature, adaptability, and feasibility of supply chain optimization decisions and ensuring that optimization decisions can truly promote the improvement of the overall operational efficiency of the supply chain.

[0071] like Figure 2 The diagram shown is a functional block diagram of a big data-driven intelligent decision-making and optimization system for the supply chain, provided in an embodiment of the present invention.

[0072] The big data-driven intelligent supply chain decision-making and optimization system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the big data-driven intelligent supply chain decision-making and optimization system 100 may include a full-link data association module 101, a standardized data slicing module 102, a bottleneck node diagnosis module 103, a mechanism analysis module 104, a decision plan generation module 105, and a comprehensive judgment module 106. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0073] In this embodiment, the functions of each module / unit are as follows: The full-link data association module 101 is used to integrate the aviation material inventory data, global positioning data, environmental perception data and settlement cycle data of the supply chain into heterogeneous data, and perform structured association on the heterogeneous data to obtain the full-link association dataset of the supply chain. The standardized data slicing module 102 is used to extract key events from the full-link associated dataset to obtain standardized data slices of the supply chain. The bottleneck node diagnosis module 103 is used to extract the operational characteristics of the standardized data slice, and based on the operational characteristics, perform network structure analysis on the supply chain to obtain the node association relationship of the supply chain, and perform supply and demand matching degree diagnosis on the node association relationship to obtain the bottleneck node of the supply chain. The mechanism analysis module 104 is used to perform mechanism analysis on the bottleneck node to obtain the key constraints of the supply chain. The decision-making plan generation module 105 is used to perform multi-path simulation and deduction on the historical data of the same period of the supply chain and the information of goods in transit based on the key constraints, so as to obtain the decision-making plan of the supply chain. The comprehensive analysis module 106 is used to comprehensively analyze the decision-making plan to obtain the optimized decision of the supply chain.

[0074] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0075] The modules described as separate components may or may not be physically separate. The components shown as modules 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 embodiment according to actual needs.

[0076] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0077] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0078] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A big data-driven intelligent decision-making and optimization method for supply chains, characterized in that, The method includes: S1. Integrate the aviation material inventory data, global positioning data, environmental perception data and settlement cycle data of the supply chain into heterogeneous data, and perform structured association on the heterogeneous data to obtain the full-link association dataset of the supply chain. S2. Extract key events from the full-link associated dataset to obtain standardized data slices of the supply chain; S3. Extract the operational characteristics of the standardized data slices, and based on the operational characteristics, perform network structure analysis on the supply chain to obtain the node relationships of the supply chain, and perform supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain. S4. Perform mechanism analysis on the bottleneck node to obtain the key constraints of the supply chain; S5. Based on the key constraints, perform multi-path simulation and deduction on the historical data of the same period of the supply chain and the information of goods in transit to obtain the decision-making plan for the supply chain. S6. Conduct a comprehensive analysis of the decision-making plan to obtain the optimized decision for the supply chain.

2. The supply chain intelligent decision-making and optimization method based on big data as described in claim 1, characterized in that, The process involves integrating supply chain data such as aircraft material inventory, global positioning data, environmental sensing data, and settlement cycle data into heterogeneous data, and then performing structured correlation on this heterogeneous data to obtain a full-link correlation dataset of the supply chain, including: Semantic parsing of fields is performed on aviation material inventory data from the procurement side, global positioning data from the transportation side, environmental perception data from the warehousing side, and settlement cycle data from the financial side to obtain the initial structured record of the supply chain. Based on the initial structured records, the task data of the supply chain is aligned across sources to obtain the task-level intermediate data of the supply chain. The task-level intermediate data is fused over time to obtain the time-series correlation data of the supply chain; Spatial registration is performed on the time-series correlated data to obtain the spatial alignment data of the supply chain; The time-series correlated data and the spatially aligned data are fused and stitched together to obtain the full-link correlated dataset of the supply chain.

3. The supply chain intelligent decision-making and optimization method based on big data as described in claim 1, characterized in that, The step of extracting key events from the entire chain-related dataset to obtain standardized data slices of the supply chain includes: By identifying the operational nodes in the full-link associated dataset, the key event types of the supply chain can be obtained. Based on the key event types, time-series fluctuations are captured in the full-link associated dataset to obtain the event timestamps of the supply chain; Based on the timestamp, the contextual traceability of the full-link associated dataset is extended to obtain dynamic data fragments of the supply chain; Field filtering is performed on the dynamic data fragments to obtain standardized data slices of the supply chain.

4. The supply chain intelligent decision-making and optimization method based on big data as described in claim 1, characterized in that, The process involves extracting operational characteristics from the standardized data slices, performing network structure analysis on the supply chain based on these characteristics, obtaining the node relationships within the supply chain, and diagnosing the supply-demand matching degree of these node relationships to identify bottleneck nodes in the supply chain. This includes: The standardized data slices are subjected to time-series analysis to obtain their dynamic operating characteristics. Based on the dynamic operation characteristics, an association topology is constructed with the nodes of the supply chain as entities and the node flow paths of the supply chain as boundaries. The flow density of the edges in the association topology is evaluated to obtain the node association relationship of the supply chain. By comparing the supply and demand scales of the associated nodes in the node association relationship, the supply and demand matching deviation of the associated nodes is obtained; Based on the supply and demand mismatch, the associated nodes are traversed and screened to obtain the bottleneck nodes of the supply chain.

5. The supply chain intelligent decision-making and optimization method based on big data as described in claim 4, characterized in that, Based on the dynamic operating characteristics, a relational topology is constructed with the nodes of the supply chain as entities and the node flow paths of the supply chain as boundaries. The flow density of the edges in the relational topology is evaluated to obtain the node association relationships of the supply chain, including: Based on the dynamic operation characteristics, business nodes and material flow paths in the supply chain are identified, and the relationship between the business nodes and the material flow paths is deduced to obtain the node and path relationship of the supply chain. Based on the relationship between the nodes and paths, an association topology is constructed with the business nodes as entities and the material flow paths as boundaries; The frequency of material interactions is statistically analyzed on the associated topology to obtain the interaction frequency feature value of the material flow path; The timeliness of the material flow path is compared by performing a flow timeliness comparison on the associated topology to obtain the timeliness stability characteristic value of the material flow path; The interaction frequency feature value and the timeliness stability feature value are coupled and evaluated to obtain the node correlation tightness of the supply chain; The node associations are sorted and categorized based on their degree of association to obtain the node association relationships in the supply chain.

6. The supply chain intelligent decision-making and optimization method based on big data as described in claim 5, characterized in that, The formula for calculating the node association density is as follows: ； in, The node association tightness, The interaction frequency feature value is... The actual circulation time of the aforementioned time-stability characteristic value. The historical average turnover time of the aforementioned time-stability characteristic value. The preset weighting coefficients, The preset weighting coefficients, It is an exponential function.

7. The supply chain intelligent decision-making and optimization method based on big data as described in claim 1, characterized in that, The mechanism analysis of the bottleneck node yields the key constraints of the supply chain, including: By tracing the upstream and downstream supply and demand relationships of the bottleneck node, the supply and demand imbalance boundary of the bottleneck node can be obtained. The operational capacity of the bottleneck node is analyzed to obtain the capacity carrying threshold of the bottleneck node. The bottleneck node is analyzed for external environment correlation to obtain the environmental constraints of the bottleneck node. By attributing and aggregating the supply and demand imbalance boundary, the capacity carrying capacity threshold, and the environmental constraints, the key constraints of the supply chain are obtained.

8. The supply chain intelligent decision-making and optimization method based on big data as described in claim 1, characterized in that, Based on the aforementioned key constraints, multi-path simulations are performed on the historical data and in-transit information of the supply chain to obtain decision-making plans for the supply chain, including: Retrieve historical data from the same period in the supply chain to obtain the historical benchmark for the supply chain; The status of in-transit goods in the supply chain is analyzed to obtain the status of in-transit goods in the supply chain. Based on the aforementioned key constraints, a perturbation injection is performed on the historical reference benchmark to obtain the first set of evolution path branches of the supply chain; Based on the aforementioned key constraints, the status of the goods in transit is prospectively extended to obtain the second set of evolution path branches of the supply chain. The first set of evolution path branches and the second set of evolution path branches are merged and deduplicated to obtain the decision plan for the supply chain.

9. The supply chain intelligent decision-making and optimization method based on big data as described in claim 1, characterized in that, The process of comprehensively evaluating the decision-making plans to obtain the optimized supply chain decision includes: The effectiveness indicators of the decision-making plan are extracted to obtain the multidimensional characteristics of the path effectiveness of the supply chain; By performing statistical analysis on the interrelationships of the efficiency dimensions of the multidimensional features, the efficiency correlation of the supply chain can be obtained. Based on the aforementioned performance correlation, scenario stress tests are conducted on the decision-making plan to obtain the scenario tolerance of the decision-making plan. Based on the scenario tolerance, the decision-making plan is analyzed at the operational micro level to obtain the resource conflict list of the decision-making plan; Based on the resource conflict list, conflict resolution is performed on the decision-making plan to obtain a feasible plan. The execution complexity of the feasible contingency plan is evaluated to obtain the execution smoothness of the decision-making contingency plan; The execution smoothness and the scenario tolerance are fused and encapsulated to obtain the optimized decision for the supply chain.

10. A big data-driven intelligent decision-making and optimization system for supply chains, characterized in that: The system, used to implement the big data-driven intelligent supply chain decision-making and optimization method as described in claim 1, comprises: The end-to-end data association module is used to integrate aviation material inventory data, global positioning data, environmental perception data and settlement cycle data of the supply chain into heterogeneous data, and to perform structured association on the heterogeneous data to obtain the end-to-end association dataset of the supply chain. The standardized data slicing module is used to extract key events from the full-link associated dataset to obtain standardized data slices of the supply chain. The bottleneck node diagnosis module is used to extract the operational characteristics of the standardized data slices, and based on the operational characteristics, perform network structure analysis on the supply chain to obtain the node relationships of the supply chain, and perform supply and demand matching degree diagnosis on the node relationships to obtain the bottleneck nodes of the supply chain. The mechanism analysis module is used to perform mechanism analysis on the bottleneck node to obtain the key constraints of the supply chain. The decision-making plan generation module is used to perform multi-path simulation and deduction based on the key constraints, historical data of the same period of the supply chain and information on goods in transit, to obtain the decision-making plan for the supply chain. The comprehensive analysis module is used to comprehensively analyze the decision-making plan to obtain the optimized decision for the supply chain.

Images