Supply chain management method and system based on data analysis processing

Abstract

The invention relates to the technical field of supply chain management, in particular to a supply chain management method and system based on data analysis processing. Comprising the following steps: S1, acquiring events of each supply chain node, and dividing the events into basic events and sudden events; the supply chain nodes are clustered according to the time proportion of the supply starting time and the supply ending time of the basic event and the sudden event at the supply chain nodes, and a clustering result is obtained; s2, based on the clustering result, extracting material flow, information flow and fund flow information in each supply chain node; and taking the circulation relation of the material flow, the information flow and the fund flow as the first attention degree of each supply chain node. According to the method, supply chain risk distribution is identified through time difference proportion clustering, adaptive adjustment and early warning prompt of a supply plan are realized by combining circulation relation deviation and exponential function dynamic analysis, and the response speed and the cooperation efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain management, and in particular to a supply chain management method and system based on data analysis and processing. Background Art

[0002] Current supply chain management relies on static forecasting models, which have significant flaws: traditional methods have difficulty dynamically distinguishing between predictable events and sudden events, resulting in delayed risk response. There is a lack of quantitative standards for the coordination of material flow, information flow, and capital flow between nodes. In particular, when non-adjacent nodes interact, insufficient data standardization leads to decision-making bias. Existing technologies do not effectively utilize the distribution characteristics of event time differences for node clustering, nor do they establish an early warning mechanism based on circulation relationship deviations, resulting in rigid resource scheduling. When anomalies occur, the system cannot quickly identify key nodes, exacerbating supply chain volatility. There is an urgent need for a solution that integrates time series clustering, dynamic evaluation of the three-flow coordination degree, and adaptive threshold management to enhance the resilience and coordination efficiency of the supply chain in complex environments. Summary of the Invention

[0003] In order to overcome the disadvantage of imbalance in the coordination of three flows under emergencies, the present invention provides a supply chain management method and system based on data analysis and processing.

[0004] The technical implementation scheme of the present invention is: a supply chain management method based on data analysis and processing, comprising the following steps: S1: Acquire events from various supply chain nodes and classify them into basic events and sudden events; cluster the supply chain nodes based on the time ratio between the start time and the end time of the basic events and sudden events at the supply chain nodes to obtain clustering results; S2: Based on the clustering results, extract the material flow, information flow and capital flow information in each supply chain node; use the circulation relationship of the material flow, information flow and capital flow as the first attention level of each supply chain node; S3: Based on the first attention level, analyzing the circulation relationship of each supply chain node to obtain an analysis result; based on the analysis result, extracting the concentration level and dispersion level of the corresponding supply chain node; S4: Based on the concentration level and the dispersion level, obtain a second level of attention of each supply chain node; and manage the supply chain based on the second level of attention.

[0005] Preferably, the acquiring of events at various supply chain nodes and the classification of the events into basic events and sudden events include: The basic event refers to an event predicted by a machine model or human experience; The sudden event refers to an event that cannot be predicted by machine models or human experience; The time difference between the start time and the end time of the basic event and the sudden event is used as the first time difference.

[0006] Preferably, clustering the supply chain nodes based on the time ratio between the supply start time and the supply end time of the basic event and the sudden event at the supply chain node to obtain the clustering result includes: The time difference between the supply start time and the supply end time of the supply chain node is used as the second time difference; Clustering is performed based on the numerical similarity of the first time difference to the second time difference to obtain a clustering result.

[0007] Preferably, extracting the material flow, information flow and capital flow information in each supply chain node based on the clustering result includes: If the supply chain nodes in the clustering result are adjacent nodes, then extracting the material flow, information flow and capital flow information of the adjacent nodes; If the supply chain nodes in the clustering result are non-adjacent nodes, normalizing the adjacent relationships between the nodes in the supply chain to obtain a first normalization result; Based on the standardization processing result, the material flow, information flow and capital flow information of the non-adjacent nodes are extracted.

[0008] Preferably, the first level of attention of each supply chain node is determined by taking the circulation relationship of the material flow, information flow, and capital flow as the first level of attention, including: The time difference between the latest supply time of the material flow and the earliest delivery time of the capital flow in adjacent nodes is used as the third time difference; The time difference between the earliest supply time of the material flow and the latest delivery time of the capital flow in the adjacent nodes is used as the fourth time difference; Taking the information sharing period of the information flow in the adjacent nodes as a reference, extracting the third time difference and the fourth time difference within the information sharing period as the supply trend change time; Based on the supply trend change time calculation method of adjacent nodes, analog logic is used to calculate the supply trend correlation of non-adjacent nodes.

[0009] Preferably, the method for calculating the supply trend change time based on adjacent nodes uses analog logic to calculate the supply trend relevance of non-adjacent nodes, including: The time difference between the earliest supply time of the material flow and the earliest delivery time of the capital flow in non-adjacent nodes is used as the fifth time difference; The time difference between the latest supply time of the material flow and the latest delivery time of the capital flow in non-adjacent nodes is used as the sixth time difference; Taking the total time length of the information sharing period of the information flow in the non-adjacent nodes as a benchmark, extracting the ratio of the sum of the fifth time difference and the sixth time difference to the total time length as the supply trend correlation; performing normalization processing on the supply trend change time and the supply trend correlation to obtain a second normalization processing result; Based on the normalization processing result, the deviation degree between the supply trend change time and the supply trend correlation is used as the first concern degree of each supply chain node.

[0010] Preferably, the analyzing the circulation relationship of each supply chain node based on the first attention level to obtain an analysis result includes: obtaining the analysis result through an analysis result formula, and the analysis result formula is as follows:

[0011] in, For analysis results, nodes and nodes Deviation measure of the circulation relationship between For nodes The first level of attention, For nodes The first level of attention, For nodes and nodes The absolute difference in the first level of attention.

[0012] Preferably, extracting the concentration and dispersion of the corresponding supply chain nodes based on the analysis results includes: Based on the analysis results, extracting deviation metrics between pairs of nodes and constructing node sequences; The analysis result value of the node sequence is used as the concentration degree of the paired supply chain nodes; The complement value of the analysis result of the node sequence is used as the dispersion degree of the paired supply chain nodes.

[0013] Preferably, obtaining a second degree of attention of each supply chain node based on the degree of concentration and the degree of dispersion; and managing the supply chain based on the second degree of attention includes: The mean value of the concentration between each supply chain node and all related nodes is used as the coordination benchmark value of the supply chain node; The average dispersion of each supply chain node and all related nodes is used as the fluctuation benchmark value of the supply chain node; The ratio of the coordination benchmark value to the fluctuation benchmark value is used as the second level of attention of the supply chain node; If the second concern level is lower than or equal to the preset threshold, the supply chain node supply plan is adjusted; If the second level of concern is higher than a preset threshold, the supply chain node monitoring status is maintained and an early warning prompt is output.

[0014] A supply chain management system based on data analysis and processing, comprising: An event clustering module is used to obtain basic events and sudden events of supply chain nodes and cluster them based on the time ratio between the start time and the end time of the supply; The three-flow extraction module is used to extract the material flow, information flow and capital flow information of adjacent nodes and non-adjacent nodes based on the clustering results, and determine the first attention level based on the circulation relationship; A relationship analysis module, configured to obtain a circulation relationship analysis result based on the first attention level through an analysis result formula, and extract the concentration level and dispersion level of the corresponding node; The supply chain management module is used to calculate the second level of concern based on the mean of the degree of concentration and the mean of the degree of dispersion, and adjust the supply plan or output early warning prompts according to the threshold.

[0015] Beneficial effects: The present invention clusters supply chain nodes by the time difference ratio between basic events and sudden events, accurately identifying risk distribution characteristics; extracts supply trend change time and correlation based on the time difference of material flow, information flow and capital flow of adjacent / non-adjacent nodes, and quantifies the first degree of concern; uses exponential functions to dynamically analyze the deviation of circulation relationships between nodes, and generates concentration and dispersion; finally, calculates the second degree of concern by the ratio of the coordination benchmark value to the fluctuation benchmark value, and automatically triggers supply plan adjustments or early warning prompts through preset thresholds, thereby realizing adaptive optimization management of the entire supply chain and significantly improving the response speed to emergencies and the efficiency of three-flow collaboration. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of the supply chain management method based on data analysis and processing of the present invention; Figure 2 This is a structural diagram of the supply chain management system based on data analysis and processing of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] Example 1: A supply chain management method based on data analysis and processing, such as Figure 1 As shown, the following steps are included: S1: Acquire events from various supply chain nodes and classify them into basic events and sudden events; cluster the supply chain nodes based on the time ratio between the start time and the end time of the basic events and sudden events at the supply chain nodes to obtain clustering results; S2: Based on the clustering results, extract the material flow, information flow and capital flow information in each supply chain node; use the circulation relationship of the material flow, information flow and capital flow as the first attention level of each supply chain node; S3: Based on the first attention level, analyzing the circulation relationship of each supply chain node to obtain an analysis result; based on the analysis result, extracting the concentration level and dispersion level of the corresponding supply chain node; S4: Based on the concentration level and the dispersion level, obtain a second level of attention of each supply chain node; and manage the supply chain based on the second level of attention.

[0019] Obtain events from each supply chain node and classify them into basic events and sudden events, including: The basic event refers to an event predicted by a machine model or human experience; The sudden event refers to an event that cannot be predicted by machine models or human experience; The time difference between the start time and the end time of the basic event and the sudden event is used as the first time difference.

[0020] It's important to note that in supply chain management, event dynamics directly impact logistics efficiency. Traditional methods struggle to distinguish between routine operations and unexpected risks. This solution, through event classification and time-quantified analysis, provides a data foundation for subsequent clustering and resource scheduling.

[0021] Event classification: basic events: predictable routine activities (such as planned procurement and regular production), which are predicted through historical data or empirical models; sudden events: unpredictable sudden disturbances (such as natural disasters and supplier bankruptcy), which are beyond the coverage of models or experience; the essence is to divide risk levels according to the predictability of events; time difference quantification, the first time difference: calculates the duration of an event from the start to the end (such as a typhoon causing a 3-day transportation interruption); this parameter converts abstract events into measurable time indicators, providing a basis for subsequent analysis of the proportion of events occupying supply chain nodes.

[0022] For example, in an automotive parts supply chain: Basic event: purchasing steel from a steel mill on the 5th of each month (predictable by human experience), lasting 2 days → first time difference = 2 days; Sudden event: a sudden operational interruption at a steel mill leads to a supply interruption (unpredictable), lasting 5 days → first time difference = 5 days; after recording these time differences, calculate their proportion of the node supply time window (e.g., 10 days) in step S1: the proportion of operational interruption events = 5 / 10 = 50%, and cluster them into one category with similar high-volatility nodes, and subsequently focus on monitoring their cash flow risks.

[0023] Clustering the supply chain nodes based on the time ratios occupied by the basic events and the sudden events at the supply chain nodes to obtain clustering results, including: The time difference between the supply start time and the supply end time of the supply chain node is used as the second time difference; Clustering is performed based on the numerical similarity of the first time difference to the second time difference to obtain a clustering result.

[0024] It should be noted that the core logic of grouping supply chain nodes by event time ratio is as follows: define the time benchmark, the second time difference: the node's own supply time window (for example, a warehouse takes 10 days from receiving an order to completing delivery); the first time difference: the actual duration of the event (basic / emergency) at the node (for example, an operational interruption causes the warehouse to stop working for 5 days); the essence is to quantify the degree to which the event squeezes the node's operating resources.

[0025] Clustering logic layer description: Feature extraction, calculation of event occupancy rate of each node: event occupancy rate = first time difference / second time difference (such as 5 / 10 = 50%); numerical similarity is calculated using distance algorithm (such as Euclidean distance): ; Example: distance between nodes A(50%) and B(45%) = =0.05, node C(10%) distance = =0.4→A and B are clustered into one category; Clustering grouping: Set a distance threshold (such as 0.1) and classify nodes with a distance ≤ the threshold into the same category (for example, if the distance between A and B is 0.05 < 0.1, they are in the same category); define grouping labels based on business needs (for example, "high volatility group": average occupancy rate within the group > 40%).

[0026] For example, the occupancy vectors of nodes X(50%), Y(60%), and Z(15%) are: [0.5, 0.6, 0.15]. Clustering process: Use the Euclidean distance formula to calculate the occupancy vectors of XY, XZ, and YZ. The calculated results of XY, XZ, and YZ are 0.1, 0.35, and 0.45, respectively. Set the distance threshold to 0.1. If the distance between X and Y is 0.1 (≤0.1), the cluster is clustered into one cluster. If the distance between Z and other nodes is greater than 0.1, the cluster is in a separate cluster. Group labeling: The average occupancy of group XY is (50% + 60%) / 2 = 55%, which is greater than 40% and is in the high volatility group. The occupancy of group Z is 15%, which is in the low volatility group.

[0027] Based on the clustering results, material flow, information flow and capital flow information in each supply chain node is extracted, including: If the supply chain nodes in the clustering result are adjacent nodes, then extracting the material flow, information flow and capital flow information of the adjacent nodes; If the supply chain nodes in the clustering result are non-adjacent nodes, normalizing the adjacent relationships between the nodes in the supply chain to obtain a first normalization result; Based on the standardization processing result, the material flow, information flow and capital flow information of the non-adjacent nodes are extracted.

[0028] It should be noted that the material flow refers to the physical movement of physical products from raw materials to consumers (forward logistics), and the reverse flow of returns / recycling. The core is the spatiotemporal transfer of physical entities; the information flow refers to the bidirectional flow of data between nodes (orders, inventory, payment status), emphasizing real-time and accuracy, and is the "nervous system" that coordinates the supply chain; the capital flow refers to financial activities (payments, credit terms) in the opposite direction of the material flow, focusing on the health and timeliness of cash flow, and reflecting the financial risks of the supply chain.

[0029] Extraction Principles and Standardization: 1. Adjacent Nodes (Direct Association): Direct extraction: For example, if an auto parts factory (node ​​A) and an assembly plant (node ​​B) are adjacent, material shipment quantities, order information, and payment receipts from A to B can be directly obtained. Advantage: Data is naturally linked, eliminating the need for conversion. 2. Non-adjacent Nodes (Indirect Association): Pain Point: For example, a raw material steel mill (node ​​C) and an auto dealership (node ​​D) have no direct interaction, but price increases at the steel mill ultimately affect the dealership's inventory.

[0030] Standardized processing method: Step 1: Construct a node relationship topology diagram and mark the hierarchy of all nodes (e.g., steel plant → parts plant → assembly plant → 4S store is a 4-level link); Step 2: Define "virtual adjacent relationships" - group cross-level nodes according to business logic (e.g., steel plant and 4S store are grouped into a "long chain response group"); Step 3: Unify data formats by group (e.g., cash flow is uniformly measured according to the "order trigger to settlement completion" cycle); Purpose: Eliminate hierarchical differences and enable comparison between non-adjacent nodes.

[0031] Example scenario: Clustering to identify high-volatility groups (warehouses X and Y); adjacent node extraction: Warehouse X is adjacent to the assembly plant → Directly obtain: Material flow: Daily number of parts shipped from X; Information flow: Order confirmation time sent by the assembly plant to X; Cash flow: Payment period for the assembly plant to X (e.g., 30 days); Non-adjacent node extraction (steel plant C and 4S dealership D): Standardization: Compress the four-level link from C to D into a "supply-sales virtual group"; Extract three flows: Material flow: Steel production capacity in C → Auto sales in D (correlations are fitted using historical data); Information flow: Raw material price fluctuation signals in C → Time lag in promotional strategy adjustments in D; Cash flow: Prepayment period in C → Installment default rate for end customers in D. Standardization solves the problem of data dimension differences between non-adjacent nodes caused by hierarchical isolation (e.g., steel mills price by ton, 4S stores sell by unit); standardization enables cross-hierarchical analysis through virtual grouping and unified measurement rules; virtual proximity relies on historical behavioral correlations (e.g., the lagged correlation model between steel mill shutdowns and 4S store inventory shortages) rather than physical connections; practical significance: If a steel mill (C) experiences a sudden operational interruption (a sudden event), the standardized model can be used to predict the cash flow risk of the 4S store (D) three months later—early warning requires shortening the customer's credit period.

[0032] The first level of attention for each supply chain node is based on the circulation relationship of the material flow, information flow, and capital flow, including: The time difference between the latest supply time of the material flow and the earliest delivery time of the capital flow in adjacent nodes is used as the third time difference; The time difference between the earliest supply time of the material flow and the latest delivery time of the capital flow in the adjacent nodes is used as the fourth time difference; Taking the information sharing period of the information flow in the adjacent nodes as a reference, extracting the third time difference and the fourth time difference within the information sharing period as the supply trend change time; Based on the supply trend change time calculation method of adjacent nodes, analog logic is used to calculate the supply trend correlation of non-adjacent nodes.

[0033] It should be noted that 1. The third time difference (latest supply vs. earliest delivery) is defined as the difference between the latest material flow supply completion time and the earliest cash flow delivery completion time between adjacent nodes; its function is to detect the aggressiveness of capital recovery (for example, customers require advance payment before parts arrive → exposing capital occupation risk); 2. The fourth time difference (earliest supply vs. latest delivery) is defined as the difference between the earliest material flow supply start time and the latest cash flow delivery completion time; its function is to identify the tendency of delayed capital payment (for example, customers delay payment after supplier shipment → reflecting cash flow pressure); design logic: the two form a symmetrical monitoring: the third time difference 1. The third and fourth time differences (↓material delay / ↑fund urgency) → early warning of supplier risks; the fourth time difference (↑material advance / ↓fund delay) → early warning of demander risks; 2. Information sharing cycle, definition: the interval for fixed exchange of key data between adjacent nodes (such as daily inventory synchronization, order confirmation per cycle); benchmark function: slicing and analyzing the third and fourth time differences according to this cycle (such as breaking down a 30-day account period into six 5-day cycles to observe trends); 3. Supply trend change time, essence: the dynamic combination of the third and fourth time differences within a single information sharing cycle; core value: quantifying the real-time fluctuations in supply and demand coordination efficiency (for example: widening time difference within a cycle = deterioration in coordination efficiency).

[0034] Example (auto parts manufacturer X and assembly plant Y). Background: Information sharing cycle = 5 days (order and payment status synchronized every 5 days). Third time lag: X's latest shipment is 3 days vs. Y's earliest payment is 1 day → time lag = 2 days. Fourth time lag: X's earliest shipment is 1 day vs. Y's latest payment is 5 days → time lag = 4 days. Supply trend change analysis: Current cycle: Third time lag (2 days) + Fourth time lag (4 days) = 6 days. Historical comparison: Previous cycle's total = 3 days (2 + 1). Trend change: Collaboration efficiency decreases by 100% ((6-3) / 3), exacerbated by Y's delayed payment. Action: Mark Y as a "high-risk cash flow node," triggering negotiations to shorten payment terms.

[0035] The third and fourth time differences are like the "double pointers of a pressure gauge" in the supply chain: the third time difference monitors the aggressiveness of the demander's capital demands (the smaller the time difference, the more exploitative the supplier); the fourth time difference monitors the security of the supplier's capital recovery (the larger the time difference, the more disadvantageous it is to the supplier); the anchoring value of the information sharing cycle: avoid fragmented analysis, bind time difference fluctuations to business rhythm (such as payment cycle, production rhythm), and ensure synchronization of trend diagnosis and decision-making.

[0036] Based on the supply trend change time calculation method of adjacent nodes, analog logic is used to calculate the supply trend correlation of non-adjacent nodes, including: The time difference between the earliest supply time of the material flow and the earliest delivery time of the capital flow in non-adjacent nodes is used as the fifth time difference; The time difference between the latest supply time of the material flow and the latest delivery time of the capital flow in non-adjacent nodes is used as the sixth time difference; Taking the total time length of the information sharing period of the information flow in the non-adjacent nodes as a benchmark, extracting the ratio of the sum of the fifth time difference and the sixth time difference to the total time length as the supply trend correlation; performing normalization processing on the supply trend change time and the supply trend correlation to obtain a second normalization processing result; Based on the normalization processing result, the deviation degree between the supply trend change time and the supply trend correlation is used as the first concern degree of each supply chain node.

[0037] It should be noted that the fifth time difference refers to the earliest supply of material flow vs. the earliest delivery of capital flow, which is used to monitor the response limit of the long chain (the speed of supply and demand coordination in an ideal scenario); the sixth time difference refers to the latest supply of material flow vs. the latest delivery of capital flow, which is used to identify the risk tolerance of the long chain (the system's ability to resist delays in the worst scenario); explanation of asymmetry: adjacent nodes focus on pressure mismatch (such as the third / fourth time difference); non-adjacent nodes focus on transmission efficiency (the fifth / sixth time difference reflects cross-level coordination consistency).

[0038] Ratio definition and function, supply trend correlation formula: ;in, is the time difference between the earliest supply time of material flow and the earliest delivery time of capital flow in non-adjacent nodes (the fifth time difference), is the time difference between the latest supply time of material flow and the latest delivery time of capital flow in non-adjacent nodes (the sixth time difference), is the cross-level data synchronization benchmark period (for example, the complete data update period from the start to the end of the supply chain, such as the quarterly synchronization period); the ratio value means, the numerator + The denominator is the total response time window of the long chain, which reflects the sum of the best and worst scenario response times between non-adjacent nodes. Represents the theoretical coordination cycle benchmark, that is, the standard period for data synchronization of non-adjacent nodes; value range interpretation: if > 1, the actual response time exceeds the benchmark, indicating low coordination efficiency; if < 1, the response is faster than the benchmark, indicating strong system resilience; standardization processing and deviation quantification, standardization method: the supply trend change time of adjacent nodes (denoted as ) and the supply trend correlation of non-adjacent nodes (denoted as ) is normalized to the interval [0,1]: ; ;in and is the minimum and maximum value of historical data, and is the minimum and maximum value of historical data; the deviation degree formula is ;in, and is the mean of the historical standardized data; the Euclidean distance is used to comprehensively evaluate the short chain pressure of the node ( ) and long chain collaboration ( ) deviates from the normal level. The larger the value, the more significant the risk.

[0039] Example (Automotive Supply Chain: Steel Plant C → 4S Store D), Background: Total Information Sharing Cycle = 90 days (quarterly data synchronization), the fifth time difference: C's earliest material supply time vs. D's earliest vehicle payment time = 30 days, the sixth time difference: C's latest material supply time vs. D's latest vehicle payment time = 120 days, supply trend correlation: correlation = (30 + 120) / 90 = 150 / 90 = 1.67 → long chain response efficiency is 67% lower than the benchmark, deviation calculation: standardization: if the historical Mean = 0.4, Mean = 0.6, current =0.5, =0.8, deviation= = =0.22→First level of concern = 0.22 (threshold 0.15), triggering an early warning; Technical advantages, risk penetration: breaking through physical proximity restrictions, capturing cross-level risk transmission (such as the impact of raw material fluctuations on terminal sales); Dynamic benchmark: The ratio mechanism automatically adapts to different business cycles to avoid absolute time misjudgment; Comprehensive decision-making: By integrating short-chain and long-chain indicators through standardized deviation, weak nodes can be accurately located; Management value: The deviation of 0.22 in the example indicates that the financial settlement terms of C→D need to be optimized (such as introducing factoring to shorten the collection period), and the credit policy of D-end customers should be monitored simultaneously.

[0040] Based on the first degree of concern, the circulation relationship of each supply chain node is analyzed to obtain an analysis result, including: obtaining the analysis result through an analysis result formula, the analysis result formula is as follows:

[0041] in, For analysis results, nodes and nodes Deviation measure of the circulation relationship between For nodes The first level of attention, For nodes The first level of attention, For nodes and nodes The absolute difference in the first level of attention.

[0042] It should be noted that the mathematical essence is: the exponential decay function, which maps the risk difference between nodes to a synergy index of [0,1]; business logic: | - |: absolute difference between the first attention level (risk deviation) of the two nodes; exponential function characteristics: when |ΔF|→0, →1 (risk symmetry → collaborative stability); when |ΔF| increases, →0 (risk imbalance → collaborative fragility); Design intention: quantify the intensity of risk transmission - the greater the difference, the easier it is for the circulation relationship to break (such as high-risk nodes dragging down low-risk nodes).

[0043] Example, scenario (first concern F value): Operational interruption warehouse X: =0.22; Assembly Plant Y: =0.15; Crisis Steel Plant C: =0.62; Synergistic deviation analysis: XY circulation relationship: = = ≈0.93 → High synergy (able to share safety stock); CY circulation relationship: = = ≈0.63 → critical risk (need to establish a financial firewall); management decision: >0.9: Maintain existing cooperation; <0.7: triggers financial instrument intervention (such as advance payment factoring).

[0044] Based on the analysis results, the concentration and dispersion of the corresponding supply chain nodes are extracted, including: Based on the analysis results, extracting deviation metrics between pairs of nodes and constructing node sequences; The analysis result value of the node sequence is used as the concentration degree of the paired supply chain nodes; The complement value of the analysis result of the node sequence is used as the dispersion degree of the paired supply chain nodes.

[0045] It should be noted that the node sequence refers to the collaborative deviation metric ( ) Arrange the node pairs in ascending order to form a full-chain relationship map and identify the collaborative efficiency gradient. The concentration level refers to the node pair Value (range [0,1]), the higher the value, the stronger the risk symmetry, suitable for resource integration and coordinated scheduling, the dispersion degree refers to 1- ( The complement of ), the higher the value, the more serious the risk imbalance, the need to isolate risks or develop hedging strategies; Design logic: concentration ( ) → measure the collaborative stability between nodes; the degree of dispersion (1- ) → Quantify the urgency of risk isolation between nodes.

[0046] Example, context: Warehouse X and Assembly Plant Y: =0.93; Steel Plant C and Warehouse X: =0.58; Steel Plant C and Assembly Plant Y: =0.63; Step decomposition: Construct node sequence (according to Ascending order): [CX(0.58), CY(0.63), XY(0.93)] → The beginning of the sequence is the most vulnerable link (CX coordination is the worst); Extraction degree index: the degree of concentration of nodes on CX ( ) is 0.58 → low concentration, dispersion (1- ) is 0.42 → highly dispersed; the degree of concentration of nodes on CY ( ) is 0.63 → moderately concentrated, with a dispersion degree of (1- ) is 0.37 → medium dispersion; the degree of concentration of nodes on XY ( ) is 0.93 → highly concentrated, with a dispersion degree (1- ) is 0.07 → low dispersion; Management decision: High concentration group (XY): concentration level 0.93 → merge inventory monitoring and share safety reserves; High dispersion group (CX): dispersion level 0.42 → establish a financial firewall (such as cash and spot).

[0047] Obtaining a second degree of attention for each supply chain node based on the degree of concentration and the degree of dispersion; and managing the supply chain based on the second degree of attention, including: The mean value of the concentration between each supply chain node and all related nodes is used as the coordination benchmark value of the supply chain node; The average dispersion of each supply chain node and all related nodes is used as the fluctuation benchmark value of the supply chain node; The ratio of the coordination benchmark value to the fluctuation benchmark value is used as the second level of attention of the supply chain node; If the second concern level is lower than or equal to the preset threshold, the supply chain node supply plan is adjusted; If the second level of concern is higher than a preset threshold, the supply chain node monitoring status is maintained and an early warning prompt is output.

[0048] It should be noted that the concentration mean refers to the node With all associated nodes of Arithmetic mean: , the formula is used to reflect the overall collaborative stability of the node (the closer the value is to 1, the more stable it is); the mean of the degree of dispersion refers to the node With all associated nodes (1- ) Arithmetic mean: , is a node The total number of associated nodes is used for arithmetic average calculation; the formula is used to quantify the overall risk transmission strength of the node (the higher the value, the more vulnerable it is); the coordination benchmark value refers to the mean of the degree of concentration → measures the collaborative efficiency of the node in integrating into the supply chain; the volatility benchmark value refers to the mean of the degree of dispersion → reveals the probability of the node triggering chain risks.

[0049] The second level of attention calculation formula is: Numerator (coordination benchmark value ↑): The stronger the node coordination capability, denominator (fluctuation benchmark value ↑): The greater the node risk volatility, ratio > 1: coordination capability suppresses risk → node security, ratio < 1: risk volatility overwhelms coordination → urgent intervention is required; threshold setting method: take the 25th percentile of this ratio during the historical normal operation period ("based on the second level of concern values ​​of all nodes in the past 12 months, sorted in ascending order, and the value at the 25th percentile is used as the threshold"; example: after sorting the historical data of 100 nodes, the 25th value is 1.2 → threshold = 1.2). Example threshold: If the historical 25th percentile = 1.2, then: < 1.2 → adjust the supply plan, > 1.2 → maintain monitoring.

[0050] For example, node data: warehouse X's associated node pair XY (0.93), XC (0.58), the mean concentration (coordination benchmark value) is (0.93 + 0.58) / 2 = 0.76, the mean dispersion (fluctuation benchmark value) is (1-0.93) + (1-0.58) = 0.07 + 0.42 = 0.49 → mean = 0.49 / 2 = 0.245; steel plant C's associated node pair CX (0.58), CY (0.63), the mean concentration (coordination benchmark value) is (0.58 + 0.63) / 2 = 0.61, the mean dispersion (fluctuation benchmark value) is (0.42 + 0.37) / 2 = 0.40; second attention level calculation: warehouse X: 0.76 / 0.25 = 3.04, steel plant C: 0.61 / 0.40 = 1.53, threshold decision ( Set threshold = 1.5): Warehouse X (3.04>1.5) → Maintain monitoring, output prompt: "X node coordinated health", Steel Plant C (1.53≈1.5) → Yellow alert, adjust plan: shorten downstream payment period, increase raw material buffer inventory; Technical advantage, dynamic balance monitoring: Coordination baseline value > 0.7 and fluctuation baseline value < 0.3 → Node is in the safe zone (such as warehouse X); When the fluctuation baseline value > the coordination baseline value (that is, the second level of concern < 1), trigger a red alert; Precise resource allocation: Steel Plant C needs to increase inventory buffer by 12% and negotiate a 15-day shortened payment period; Network-level resilience: When the average second level of concern of the entire chain is < 1.3, initiate a global risk hedging strategy; Key conclusion: This mechanism converges complex node relationships into a single action indicator (second level of concern), achieving closed-loop management from diagnosis to decision-making.

[0051] Example 2: Based on Example 1, a supply chain management system based on data analysis and processing, such as Figure 2 Shown, including: An event clustering module is used to obtain basic events and sudden events of supply chain nodes and cluster them based on the time ratio between the start time and the end time of the supply; The three-flow extraction module is used to extract the material flow, information flow and capital flow information of adjacent nodes and non-adjacent nodes based on the clustering results, and determine the first attention level based on the circulation relationship; A relationship analysis module, configured to obtain a circulation relationship analysis result based on the first attention level through an analysis result formula, and extract the concentration level and dispersion level of the corresponding node; The supply chain management module is used to calculate the second level of concern based on the mean of the degree of concentration and the mean of the degree of dispersion, and adjust the supply plan or output early warning prompts according to the threshold.

[0052] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A supply chain management method based on data analysis and processing, characterized in that: The following steps are involved: S1: Obtain events at each supply chain node and classify the events into basic events and sudden events; Clustering the supply chain nodes based on the time ratios occupied by the basic events and the sudden events at the supply chain nodes' start and end times to obtain clustering results; S2: Based on the clustering results, extract the material flow, information flow and capital flow information in each supply chain node; use the circulation relationship of the material flow, information flow and capital flow as the first attention level of each supply chain node; S3: Analyze the circulation relationship of each supply chain node based on the first attention level to obtain an analysis result; Based on the analysis results, extract the concentration and dispersion of the corresponding supply chain nodes; S4: obtaining a second degree of attention for each supply chain node based on the degree of concentration and the degree of dispersion; Based on the second level of concern, the supply chain is managed.

2. A supply chain management method based on data analysis and processing as claimed in claim 1, characterized in that: The events of each supply chain node are obtained and divided into basic events and sudden events, including: The basic event refers to an event predicted by a machine model or human experience; The sudden event refers to an event that cannot be predicted by machine models or human experience; The time difference between the start time and the end time of the basic event and the sudden event is used as the first time difference.

3. The supply chain management method based on data analysis and processing as claimed in claim 2, characterized in that: Clustering the supply chain nodes based on the time ratios occupied by the basic events and the sudden events at the supply chain nodes to obtain clustering results includes: The time difference between the supply start time and the supply end time of the supply chain node is used as the second time difference; Clustering is performed based on the numerical similarity of the first time difference to the second time difference to obtain a clustering result.

4. The supply chain management method based on data analysis and processing according to claim 1, characterized in that: The extracting of material flow, information flow and capital flow information in each supply chain node based on the clustering result includes: If the supply chain nodes in the clustering results are adjacent nodes, extracting the material flow, information flow, and capital flow information of the adjacent nodes; If the supply chain nodes in the clustering result are non-adjacent nodes, normalizing the adjacent relationships between the nodes in the supply chain to obtain a first normalization result; Based on the standardization processing result, the material flow, information flow and capital flow information of the non-adjacent nodes are extracted.

5. The supply chain management method based on data analysis and processing as claimed in claim 1, characterized in that: The first level of attention of each supply chain node is based on the circulation relationship of the material flow, information flow, and capital flow, including: The time difference between the latest supply time of the material flow and the earliest delivery time of the capital flow in adjacent nodes is used as the third time difference; The time difference between the earliest supply time of the material flow and the latest delivery time of the capital flow in the adjacent nodes is used as the fourth time difference; Taking the information sharing period of the information flow in the adjacent nodes as a reference, extracting the third time difference and the fourth time difference within the information sharing period as the supply trend change time; Based on the supply trend change time calculation method of adjacent nodes, analog logic is used to calculate the supply trend correlation of non-adjacent nodes.

6. The supply chain management method based on data analysis and processing as claimed in claim 5, characterized in that: The method for calculating the supply trend change time based on adjacent nodes uses analog logic to calculate the supply trend correlation of non-adjacent nodes, including: The time difference between the earliest supply time of the material flow and the earliest delivery time of the capital flow in non-adjacent nodes is used as the fifth time difference; The time difference between the latest supply time of the material flow and the latest delivery time of the capital flow in non-adjacent nodes is used as the sixth time difference; Taking the total time length of the information sharing period of the information flow in the non-adjacent nodes as a benchmark, extracting the ratio of the sum of the fifth time difference and the sixth time difference to the total time length as the supply trend correlation; performing normalization processing on the supply trend change time and the supply trend correlation to obtain a second normalization processing result; Based on the normalization processing result, the deviation degree between the supply trend change time and the supply trend correlation is used as the first concern degree of each supply chain node.

7. The supply chain management method based on data analysis and processing according to claim 1, characterized in that: The step of analyzing the circulation relationship of each supply chain node based on the first degree of attention to obtain an analysis result includes: obtaining the analysis result through an analysis result formula, wherein the analysis result formula is as follows: in, To analyze the results, the node and nodes Deviation measure of the circulation relationship between For nodes The first level of attention, For nodes The first level of attention, For nodes and nodes The absolute difference in the first level of attention.

8. The supply chain management method based on data analysis and processing as claimed in claim 7, characterized in that: Extracting the concentration and dispersion of the supply chain nodes based on the analysis results includes: Based on the analysis results, extracting deviation metrics between pairs of nodes and constructing node sequences; The analysis result value of the node sequence is used as the concentration degree of the paired supply chain nodes; The complement value of the analysis result of the node sequence is used as the dispersion degree of the paired supply chain nodes.

9. The supply chain management method based on data analysis and processing according to claim 1, characterized in that: obtaining a second degree of attention of each supply chain node based on the degree of concentration and the degree of dispersion; Based on the second level of concern, supply chain management includes: The mean value of the concentration between each supply chain node and all related nodes is used as the coordination benchmark value of the supply chain node; The average dispersion of each supply chain node and all related nodes is used as the fluctuation benchmark value of the supply chain node; The ratio of the coordination benchmark value to the fluctuation benchmark value is used as the second level of attention of the supply chain node; If the second concern level is lower than or equal to the preset threshold, the supply chain node supply plan is adjusted; If the second level of concern is higher than a preset threshold, the supply chain node monitoring status is maintained and an early warning prompt is output.

10. A supply chain management system based on data analysis and processing, used to implement the supply chain management method based on data analysis and processing as described in any one of claims 1 to 9, characterized in that: include: An event clustering module is used to obtain basic events and sudden events of supply chain nodes and cluster them based on the time ratio between the start time and the end time of the supply; The three-flow extraction module is used to extract the material flow, information flow and capital flow information of adjacent nodes and non-adjacent nodes based on the clustering results, and determine the first attention level based on the circulation relationship; A relationship analysis module, configured to obtain a circulation relationship analysis result based on the first attention level through an analysis result formula, and extract the concentration level and dispersion level of the corresponding node; The supply chain management module is used to calculate the second level of concern based on the mean of the degree of concentration and the mean of the degree of dispersion, and adjust the supply plan or output early warning prompts according to the threshold.

Images