A supply chain inventory dynamic management method and system based on an industrial internet platform

Abstract

The application discloses a kind of based on industrial internet platform's supply chain inventory dynamic management method and system, the method includes: obtaining and freight state information related to natural sudden event of supply chain operation, and freight state information is carried out uncertainty quantification evaluation and correction, constructs material running external environment information;Material running external environment information is carried out data stream feature analysis and physical mapping consistency evaluation, obtain target supplier state determination result;According to target supplier state determination result, carry out reliability weight dynamic adjustment, and according to weight dynamic adjustment result, inventory data availability weight is valued, realizes supply chain inventory dynamic management.The present application can improve the risk identification and response capability of industrial internet platform when facing external data source interruption.The present application as a kind of based on industrial internet platform's supply chain inventory dynamic management method and system, can be widely applied in supply chain inventory dynamic management technical field.

Description

Technical Field

[0001] This invention relates to the field of dynamic supply chain inventory management technology, and in particular to a method and system for dynamic supply chain inventory management based on an industrial internet platform. Background Technology

[0002] In industrial internet platforms deployed by large manufacturing enterprises, their supply chain inventory dynamic management modules heavily rely on real-time material production progress, inventory status, and transportation information obtained from external entities such as upstream suppliers and logistics carriers. However, when sudden external events (such as natural disasters or regional communication disruptions) cause interruptions or severe delays in the transmission of these critical external data sources, existing platforms often struggle to effectively identify and quantify the "failure" or "reliability degradation" of these external data streams. When the platform fails to receive new real-time data updates, its automated decision-making engine tends to continue using old data from the "last known normal state" for inventory assessment and cross-regional transfer instructions. This blind reliance on faulty data means that when faced with fundamental disruptions in the upstream supply chain, the platform not only fails to provide timely warnings but also accelerates the generation and execution of erroneous automated decisions based on outdated or incomplete information, such as initiating invalid material transfers. This rapidly amplifies localized data problems into global production stagnation and significant economic losses. Summary of the Invention

[0003] To address the aforementioned technical problems, the purpose of this invention is to provide a method and system for dynamic supply chain inventory management based on an industrial internet platform, which can enhance the risk identification and response capabilities of the industrial internet platform when facing external data source interruptions.

[0004] The first technical solution adopted in this invention is: a method for dynamic management of supply chain inventory based on an industrial internet platform, comprising the following steps: Acquire information on natural emergencies and freight status related to supply chain operations, and conduct uncertainty quantification assessment and correction of freight status information to construct information on the external environment of material movement; Data flow feature analysis and physical mapping consistency assessment are performed on the external environment information of material operation to obtain the target supplier status determination result; Based on the target supplier status assessment, the reliability weight is dynamically adjusted, and the validity weight of the inventory data is assigned according to the dynamic adjustment result, thereby realizing dynamic management of supply chain inventory.

[0005] Furthermore, the step of acquiring information on natural emergencies and freight status related to supply chain operations, and performing uncertainty quantification assessment and correction on the freight status information to construct external environment information for material operations specifically includes: Acquire information on natural emergencies and freight status related to supply chain operations. Natural emergencies include natural disasters, regional power outages, communication network failures, or traffic control. Freight status information includes material transport speed and estimated arrival time. Uncertainty quantification and correction are performed on freight status information to obtain corrected freight status information; By cross-integrating natural emergencies with the corrected freight status information, we can construct external environment information for material operation.

[0006] Furthermore, the step of performing uncertainty quantification assessment and correction on the freight status information to obtain the corrected freight status information specifically includes: The multi-source physical world information of the carrier corresponding to the freight status information is obtained. The multi-source physical world information includes traffic flow data, road monitoring images, traffic flow sensing speed data and comparison frequency. The comparison frequency represents the comparison frequency between the material transportation speed in the freight status information and the traffic flow sensing speed data in the multi-source physical world information within a preset time range. The deviation between the material transport speed in the freight status information and the multi-source physical world information is quantified and tracked to obtain the initial deviation value and the deviation duration. Uncertainty scores are calculated based on the initial deviation value and the duration of the deviation. The estimated arrival time in the freight status information is then corrected based on the results of the uncertainty score calculation to obtain the corrected freight status information.

[0007] Furthermore, the expression for calculating the uncertainty score is as follows:

[0008] In the above formula, Represents the uncertainty score. This represents the initial deviation value. Indicates the duration of the deviation. Indicates the preset time unit. This indicates a preset, persistent impact factor.

[0009] Furthermore, the step of performing data flow feature analysis and physical mapping consistency assessment on the external environment information of material operation to obtain the target supplier status determination result specifically includes: Perform data flow characteristic analysis on information about the external environment of material operation; If the data stream characteristics of the material operation external environment information are that the data stream is in a stopped state within a preset time period or the data stream remains unchanged within a preset time period, then the target supplier is determined to be in a failed state, and the first determination result is output. If the data flow characteristics of the material operation external environment information are that there are sporadic information inputs, then a physical mapping consistency assessment is performed on the target supplier, and a first judgment result or a second judgment result is output based on the assessment results. The first or second determination result of the output is used as the determination result of the target supplier status.

[0010] Furthermore, the step of performing a physical mapping consistency assessment on the target supplier and outputting a first or second judgment result based on the assessment result if the data stream characteristics of the material's external environment information show sporadic information input specifically include: If the data stream characteristics of the material operation external environment information are that there are sporadic information inputs, the sporadic information can be identified by keyword body recognition through natural language processing to obtain sporadic key fragment information. Collect physical anchor information corresponding to scattered key fragment information, including industrial electricity consumption, satellite remote sensing data, raw material logistics orders, and geographical location information; A physical mapping consistency score is calculated between scattered key fragment information and physical anchor point information to obtain the physical mapping consistency score calculation result; Set a threshold for valid information and compare it with the physical mapping consistency score calculation results; If the physical mapping consistency score is less than the effective information threshold, the scattered information is considered invalid information, the target supplier is determined to be in a failed state, and the first determination result is output. If the physical mapping consistency score is greater than the effective information threshold, then the scattered information is considered as information to be considered, and the second judgment result is output.

[0011] Furthermore, the expression for calculating the physical mapping consistency score is as follows:

[0012] In the above formula, This represents the result of the physical mapping consistency score calculation. Indicates the first Consistency index between physical anchor points and scattered information Indicates the first The importance weight of each physical anchor point.

[0013] Furthermore, the step of dynamically adjusting reliability weights based on the target supplier status determination results, and assigning values ​​to inventory data validity weights based on the dynamic weight adjustment results to achieve dynamic supply chain inventory management, specifically includes: Based on the target supplier status assessment results, the reliability weight is dynamically adjusted; If the target supplier status determination result is the first determination result, then the inventory data validity weight corresponding to the target supplier is set to the first preset value, and the adjustment result of the first inventory data validity weight is obtained; If the target supplier status determination result is the second determination result, then the reliability weight is dynamically adjusted based on the physical mapping consistency score calculation result, and the inventory data validity weight corresponding to the target supplier is obtained by updating the inventory data validity weight value, and the second inventory data validity weight adjustment result is obtained. The updated inventory data validity weight value is greater than the first preset value. The final inventory data validity weight adjustment result is obtained by adjusting either the first or second inventory data validity weight based on the output, thus achieving dynamic inventory management in the supply chain.

[0014] Furthermore, the specific expression for the dynamic adjustment of the reliability weight is as follows:

[0015] In the above formula, This represents the validity weight value of the updated inventory data. This represents the first preset value. Indicates the physical mapping consistency score. Indicates the threshold of valid information. This represents the midpoint between the effective information threshold and the first preset value.

[0016] The second technical solution adopted in this invention is: a dynamic supply chain inventory management system based on an industrial internet platform, comprising: The first module is used to acquire information on natural emergencies and freight status related to supply chain operations, and to perform uncertainty quantification assessment and correction on freight status information to construct information on the external environment of material operation. The second module is used to perform data flow feature analysis and physical mapping consistency assessment on the external environment information of material operation to obtain the target supplier status determination result. The third module is used to dynamically adjust the reliability weight based on the target supplier status determination result, and assign a value to the validity weight of the inventory data based on the dynamic weight adjustment result, so as to realize dynamic management of supply chain inventory.

[0017] The beneficial effects of the method and system of this invention are as follows: This invention acquires information on natural emergencies and freight status related to supply chain operations, and performs uncertainty quantification assessment and correction on freight status information to construct external environment information for material operation. By quantifying deviations, it enhances the robustness of decision-making, enabling the platform to make responsible and flexible plans even when information is not entirely accurate. It can build buffers and adjustment space in advance, avoiding global production chaos caused by a single prediction failure at critical moments. Furthermore, it performs data flow feature analysis and physical mapping consistency assessment on the external environment information for material operation to obtain the target supplier status judgment result. Through rigorous physical mapping verification, it minimizes the decision-making risks brought by "partially recovered" data. This allows the platform to provide a more robust and realistic judgment when information is extremely uncertain, thereby avoiding the acceleration and amplification of erroneous automated decisions in information blind spots. Finally, based on the target supplier status judgment result, it dynamically adjusts the reliability weight and assigns a value to the validity weight of inventory data based on the dynamic weight adjustment result, realizing dynamic management of supply chain inventory. This significantly improves the risk identification and response capabilities of the industrial internet platform when facing external data source interruptions. Through multi-dimensional cross-validation of information, it ensures that the platform can accurately identify the failure status of data sources, rather than misjudging them as temporary communication failures. This enables the platform to dynamically adjust the reliability level of internal inventory data, promptly suspend erroneous automated decisions, and force manual intervention, thereby effectively avoiding ineffective allocations, production stoppages, and huge economic losses caused by information asymmetry, and ensuring the resilience and production continuity of the enterprise's supply chain. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the steps of a dynamic supply chain inventory management method based on an industrial internet platform according to the present invention. Figure 2 This is a structural block diagram of a supply chain inventory dynamic management system based on an industrial internet platform according to the present invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. The step numbers in the following embodiments are only for ease of explanation and do not limit the order of the steps. The execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art.

[0020] First, it's important to note that in the operations of modern large-scale manufacturing enterprises, dynamic supply chain inventory management is crucial for ensuring smooth production and market responsiveness. With the widespread application of industrial internet platforms, enterprises expect to optimize material flow and inventory levels through real-time data collection and automated decision-making. However, when the external data sources these platforms rely on, especially the information flow from key upstream suppliers, are interrupted or fail due to unforeseen events, ensuring that the platform can accurately identify this data failure and avoid making erroneous automated decisions based on outdated information becomes a pressing challenge.

[0021] Based on this, this invention establishes a multi-dimensional, proactive mechanism for perceiving external environmental information and assessing the reliability of internal data in material handling. It transcends traditional single-source monitoring by integrating "heartbeat" detection, data flow behavior pattern analysis, and "contextual awareness" information from public news and logistics carriers to comprehensively and in real-time cross-validate the health status of external data sources. When the core supplier's data flow is interrupted due to unforeseen events, the platform no longer blindly uses old data or simply waits. Instead, it intelligently identifies the "failure" state of the data source and dynamically adjusts the reliability level of internal inventory data, marking it as "high-risk." This provides accurate and timely information for subsequent automated decision-making adjustments and manual intervention, avoiding erroneous decisions caused by information asymmetry.

[0022] Reference Figure 1 This invention provides a method for dynamic supply chain inventory management based on an industrial internet platform, the method comprising the following steps: S100: Acquire information on natural emergencies and freight status related to supply chain operations, and conduct uncertainty quantification assessment and correction of freight status information to construct information on the external environment of material operation; S110. Obtain information on natural emergencies and freight status related to supply chain operations. The natural emergencies include natural disasters, regional power outages, communication network failures, or traffic control. The freight status information includes material transport speed and estimated arrival time. In this embodiment, the industrial internet platform continuously operates an "external environment information collector," which proactively acquires external information related to supply chain operations from multiple public channels. Specifically, the collector captures publicly available news reports and related discussions on social media globally in real time, focusing on unforeseen events that may affect the stability of the supply chain, such as natural disasters (e.g., earthquakes, floods, typhoons), regional power supply disruptions, communication network failures, or traffic control measures.

[0023] To achieve this, the collector can integrate multiple data interfaces. For example, for news reporting, it can subscribe to RSS feeds from mainstream news organizations, or use natural language processing technology to monitor and extract information in real time from specific keywords (such as "power outage," "disaster," "traffic disruption," and "supplier name"). For social media, it can utilize publicly available APIs to perform trend analysis on trending topics in specific regions or related to suppliers, identifying abnormal fluctuations.

[0024] Meanwhile, the collector connects to the systems of multiple partner logistics carriers through a pre-defined application programming interface (API) to obtain the real-time geographical location, transportation status (e.g., shipped, in transit, arrived at transit station), and estimated arrival time of all materials in transit. It pays special attention to any abnormal transportation delays or route changes. For example, if a shipment originally expected to arrive in three days shows its status as having been stuck at a transit station for more than 24 hours without being updated, or if the transportation route suddenly deviates from the regular path, these will be marked as abnormal.

[0025] For example, when the platform needs to monitor a chip supplier located in a certain area, the external environment information collector will continuously monitor earthquake warnings issued by the local meteorological agency, local news media reports on power supply, and updates from international logistics companies (such as DHL and FedEx) on the status of shipments originating from that location and destined for the rest of the world. These information streams are refreshed and analyzed every minute to ensure that the platform can detect any changes in the external environment in real time.

[0026] S120. Perform uncertainty quantification assessment and correction on the freight status information to obtain the corrected freight status information; First, it's important to clarify that during a routine material transportation process, after a critical batch of processor chips was shipped from the upstream supplier, the logistics carrier's system updated its initial transportation status and estimated arrival time. The platform, through its external information collection module, promptly acquired this preliminary transportation information and incorporated it into the in-transit material tracking system. Subsequently, while passing through a significant inter-regional land transportation hub, this batch of materials encountered a sudden, localized, non-public, and brief temporary traffic control measure, such as emergency equipment maintenance within the hub lasting approximately four hours and affecting only specific lanes. Because this control measure was not a major public event and its duration was limited, the logistics carrier's automated tracking system failed to detect this lane-specific physical delay in the short term. Its internal update logic, likely based on pre-defined route planning and average travel speed, may have made only minor adjustments to the estimated arrival time, or even no adjustments at all, because its data update frequency or event classification mechanism failed to recognize this fine-grained, transient physical interruption.

[0027] The industrial internet platform's "external environment information collector" queries the logistics carrier's system at a preset frequency, such as every 15 minutes. It receives updated transportation data, but this data still doesn't fully reflect the actual physical delays. Therefore, the platform's internal assessment of the "in transit" status and "estimated arrival time" for this batch of materials accumulates discrepancies with the actual physical reality. For example, based on the carrier's report, the platform might believe the estimated arrival time is only 30 minutes later than planned, but in reality, due to physical delays, the actual arrival time will be several hours later. At this critical moment, the destination production base needs to urgently adjust its production plan due to a sudden change in market demand, making the timeframe for the processor chips even more pressing. Based on its internal "estimated arrival time" (updated based on inaccurate logistics data), the platform determines that the materials can still arrive before the new, urgent timeframe, and thus makes further automated production scheduling and material usage planning based on this information. For example, the platform instructs the production line to prepare other supporting materials in advance and adjusts the operating parameters of relevant equipment to ensure immediate production upon the arrival of this batch of chips. However, because the logistics data collected by the external environment information collector fails to accurately reflect the actual physical delays, the automated decisions made by the platform will ultimately lead to chaos in production planning and waste of resources.

[0028] In summary, in the dynamic management of supply chain inventory on the industrial internet platform of large manufacturing enterprises, when the "external environment information collector" obtains the transportation status of materials in transit from the cooperative logistics carrier system, how can it effectively identify and quantify the subtle cumulative deviations between the transportation status reported by the carrier system and sudden, localized, non-public, and short-term physical delay events in the physical world due to the carrier system's data update frequency, classification granularity, or event perception limitations? This is to prevent the platform from generating and executing automated production scheduling and material usage planning that are out of touch with physical reality based on such incomplete and inaccurate external context information, thereby affecting the overall efficiency and responsiveness of the supply chain.

[0029] In this embodiment, the industrial internet platform continuously runs a "physical state comparison module." This module acts like a rigorous inspector, frequently comparing the in-transit material transportation status (e.g., location, speed, estimated arrival time) reported by the logistics carrier's system with multi-source physical world information independent of the carrier. Specifically, this includes: S121. Obtain multi-source physical world information of the carrier corresponding to the freight status information. The multi-source physical world information includes traffic flow data, road monitoring images, traffic flow sensing speed data and comparison frequency. The comparison frequency represents the comparison frequency between the material transportation speed in the freight status information and the traffic flow sensing speed data in the multi-source physical world information within a preset time range. S122. Quantify and track the deviation between the material transport speed in the freight status information and the multi-source physical world information to obtain the initial deviation value and the deviation duration. S123. Calculate the uncertainty score based on the initial deviation value and the duration of the deviation, and correct the estimated arrival time in the freight status information based on the result of the uncertainty score calculation to obtain the corrected freight status information.

[0030] In this embodiment, traffic flow data fusion is first performed. The platform connects to the urban traffic management system on key logistics hubs or transportation routes through a preset digital interface to obtain real-time traffic flow sensor data for that area. For example, it can obtain data such as vehicle flow rate and average vehicle speed updated every minute. The physical state comparison module geospatially matches the material location reported by the carrier with this traffic data. If the carrier reports that the material is traveling at high speed on a certain road segment, but the traffic flow sensor shows that the vehicle speed on that road segment is less than 10 kilometers per hour and the traffic flow is saturated, this is recorded as a consistency deviation.

[0031] The platform then performs road monitoring image analysis, integrating components to analyze video streams from public road surveillance cameras at key transportation nodes or sections. This component utilizes image processing techniques, such as vehicle detection and tracking algorithms based on the OpenCV library, to analyze vehicle density, speed, and stationary states in video frames in real time. A physical state comparison module compares these analysis results with the material status reported by the carrier. For example, if the carrier reports materials passing through an intersection, but the monitoring footage shows vehicles stationary at that intersection for an extended period, this is recorded as a consistency deviation. To protect privacy, the video stream can be anonymized, extracting only vehicle motion features.

[0032] Furthermore, the platform correlates environmental sensor data. It acquires environmental sensor data along key transportation routes via digital interfaces, such as real-time wind speed, rainfall, and visibility data provided by weather stations. The physical state comparison module then correlates this environmental data with the transportation status reported by the carrier. For example, if the carrier reports that materials maintain normal speed despite severe weather (such as heavy rain or fog), but environmental sensor data shows extremely low visibility in the area, this is recorded as a consistency deviation.

[0033] Finally, the comparison frequency is set. For high-value, time-sensitive key materials, the physical state comparison module can be set to perform a comparison every minute to ensure timely detection of subtle deviations.

[0034] Furthermore, when the physical state comparison module detects a discrepancy between the carrier's report and physical world information, the platform activates an "uncertainty quantification module." This module does not directly correct the estimated arrival time, but rather dynamically calculates an "estimated arrival time uncertainty score" based on the degree, duration, and historical deviation patterns. This score is then used to adjust the "confidence interval width" associated with the estimated arrival time in real time. Specifically, this includes: 1) Deviation Quantification: The uncertainty quantification module calculates an initial deviation value based on the type and severity of the deviation reported by the physical state comparison module. For example, if the difference between the speed reported by the carrier and the actual speed measured by the traffic flow sensor exceeds 30%, the initial deviation value is set to 0.3. If the monitoring image shows that the material has been stagnant for more than 15 minutes, the initial deviation value is set to 0.5.

[0035] 2) Duration Impact: The uncertainty quantification module tracks the duration of the deviation. The longer the deviation lasts, the greater its impact on the uncertainty score. For example, the uncertainty score increases by 0.1 for every 10 minutes.

[0036] 3) Calculation of uncertainty score: The uncertainty score (U) can be calculated using the following formula:

[0038] In the above formula, Represents the uncertainty score. This represents the initial deviation value. Indicates the duration of the deviation. Indicates the preset time unit. This indicates a preset, persistent impact factor.

[0039] For example, if the preset time unit is 10 minutes and the duration of influence is 0.1, and the initial deviation value is 0.3, and the deviation lasts for 30 minutes, then U = 0.3 + (30 / 10) 0.1 = 0.3 + 0.3 = 0.6.

[0040] 4) Confidence Interval Width Adjustment: The platform dynamically adjusts the confidence interval width (CIW) of the Estimated Time of Arrival (ETA) based on the calculated uncertainty score U. The initial ETA may have a default CIW0 (e.g., plus or minus 15 minutes). The new CIW is calculated as follows: CIW = CIW0 (1 + U × magnification factor).

[0041] For example, if CIW0 is 30 minutes (±15 minutes), the magnification factor is 2. When U=0.6, CIW = 30 (1 + 0.6×2) = 30 (1 + 1.2) = 30×2.2 = 66 minutes. This means that the uncertainty of the estimated arrival time widens from ±15 minutes to ±33 minutes.

[0042] 5) Threshold Judgment: The platform presets a "high-risk uncertainty threshold" (e.g., when U > 0.5 or CIW > 60 minutes). Once the calculated uncertainty score or confidence interval width exceeds this threshold, the platform determines that it has entered a "high-risk situation".

[0043] Furthermore, when the platform determines that a "high-risk situation" has been entered, it immediately activates a "risk situation driver." This driver does not attempt to directly correct the estimated arrival time, but instead matches the most relevant response from a pre-established "risk situation library" based on the currently detected level of uncertainty and type of physical deviation. Specifically, this includes: 1) Risk Situation Library: The platform maintains a structured "risk scenario library," which contains various predefined risk scenarios and their corresponding triggering conditions and response plans. Each scenario is described by a set of features, for example: Scenario 1: A critical chip is stuck at a transportation hub, with an uncertainty score U > 0.5, and the delay is expected to exceed 2 hours.

[0044] Scenario 2: Secondary critical components are stuck on ordinary road sections, with an uncertainty score U > 0.3, and the expected delay is more than 1 hour.

[0045] Scenario 3: High-value raw materials are transported in severe weather, with an uncertainty score U > 0.4 and visibility less than 50 meters.

[0046] 2) Scenario Matching Logic: The risk scenario driver compares the current material's "uncertainty score," "physical deviation type" (e.g., traffic congestion, equipment maintenance, severe weather), "material importance" (e.g., critical chips, secondary critical components), and "expected delay time range" with the scenario features in the risk scenario database in real time. The matching algorithm can employ rule-based reasoning or fuzzy matching. For example, if the current scenario has the highest feature match with "Scenario 1," then the contingency plan corresponding to "Scenario 1" is selected.

[0047] 3) Driver activation: Once a match is successful, the risk scenario driver will immediately activate the multi-level production planning adaptive mechanism corresponding to that scenario.

[0048] Finally, the multi-level adaptive production planning mechanism is triggered. Based on the contingency plans matched by the risk scenario driver, the platform will trigger a series of preset, multi-level adaptive production planning actions to cope with the high uncertainty of material arrival times. Specifically, these include: 1) Generation of Alternative Material Solutions: If the matched scenario involves a severe delay in critical materials (such as processor chips), the platform will automatically query the "Alternative Material List". For example, if chip model A is currently delayed, the platform will check if there is inventory of compatible chip models B or C, and generate a report that includes alternative material models, available quantities, performance differences, and cost impacts for manual decision-making.

[0049] 2) Dynamically adjust production line priorities: The platform will dynamically adjust the priorities of production lines based on the impact of material delays on different product production lines. For example, if the delayed chip is a critical material for a high-end smartphone production line, and that production line is extremely time-sensitive, the platform may temporarily lower its priority while increasing the priority of other product production lines that have a lower dependence on the delayed material or have sufficient local inventory, in order to maximize overall production efficiency.

[0050] 3) Activate the cross-base material transfer plan: The platform will query the inventory of similar materials in other production bases or warehousing centers around the world. If there is surplus inventory and the transfer time window allows, the platform will generate an "emergency transfer proposal" that includes the transfer quantity, transportation method (e.g., air freight, land freight), estimated arrival time, and cost estimate, along with a risk warning, and await manual approval.

[0051] 4) Mandatory Human Intervention and Physical Inventory: Regardless of the automated recommendations above, upon entering a high-risk scenario, the platform will automatically send an emergency alert (e.g., via corporate instant messaging or SMS) to the purchasing manager, supply chain director, and relevant warehouse center managers, requiring them to immediately intervene and conduct a physical inventory check of the relevant materials. Based on the physical mapping consistency assessment report and uncertainty score provided by the platform, a human assessment will determine whether to initiate alternative supplier procurement, adjust production plans, or take other contingency measures. The platform will provide an interactive interface displaying all relevant data, scenario matching results, and suggested contingency plans to assist human decision-making.

[0052] Based on this, when the platform needs to manage the inventory of critical materials at production bases and warehousing centers around the world, it continuously cross-checks the transportation information such as material location and speed reported by logistics partners with objective information from independent channels. These independent channels include real-time traffic flow data provided by urban traffic management systems, vehicle density and speed analyzed by road surveillance cameras, and environmental data from weather stations along the route. The platform acts like a meticulous observer; if it detects that a logistics partner reports materials moving at high speed, but traffic flow sensors show severe congestion on that section of road, or that monitoring footage shows materials remaining stationary for an extended period, it immediately recognizes a discrepancy between reported and actual data.

[0053] Once such inconsistencies are detected, the platform will not directly attempt to correct the estimated arrival time reported by its logistics partners, as it understands that precise predictions may be unrealistic in some cases. Instead, it dynamically calculates an "uncertainty score for estimated arrival time" based on the degree and duration of the inconsistency, thereby widening the range of risks included in the estimated arrival time. For example, if the original estimated arrival time was 3 PM with a margin of error of plus or minus 15 minutes, now, due to physical delays, the uncertainty score increases, and the margin of error may widen to plus or minus 2 hours.

[0054] When this uncertainty score reaches a preset warning threshold, the platform immediately activates a "risk scenario driver." This driver acts like an emergency command center; instead of speculating on when materials will arrive, it quickly matches the most suitable response from a pre-established "risk scenario library" based on the current level of uncertainty, the type of physical deviation, and the importance of the materials. For example, if the matched scenario is "critical chips are severely delayed at important transportation hubs," the driver will immediately trigger a series of preset adaptive actions: it might suggest using chips from alternative suppliers, dynamically adjust production line priorities, initiate an assessment of cross-site material allocation, and forcibly send an emergency alert to supply chain managers, requiring them to intervene immediately for manual confirmation and decision-making.

[0055] S130. Cross-integrate natural emergencies with the corrected freight status information to construct external environment information for material operation.

[0056] This embodiment ensures that only information truly reflecting the state of the physical world is used to guide subsequent actions by adhering to verification methods derived from objective phenomena in the physical world. It recognizes that during periods of subtle deviations in logistics information, the most important thing is to quickly identify such uncertainties and activate a comprehensive risk response mechanism. Instead of attempting to eliminate deviations in logistics information, it enhances the robustness of decision-making by quantifying these deviations, enabling the platform to make responsible and flexible plans even when information is not entirely accurate. This allows the platform to proactively build buffers and adjustment spaces when facing uncertainties arising from subtle deviations between logistics partner system reports and the physical world during the transportation of critical, high-value materials. This avoids global production chaos caused by a single failed prediction at critical moments, thereby ensuring the continuity and market competitiveness of enterprises in the production of high-tech products.

[0057] S200: Perform data flow characteristic analysis and physical mapping consistency assessment on the external environment information of material operation to obtain the target supplier status determination result; In this embodiment, the solution uses multi-dimensional cross-validation to accurately identify the true state of "data failure," rather than simply "communication failure." It can distinguish between a brief network fluctuation and a systemic collapse of the upstream supplier, thus avoiding being misled by outdated information at critical moments. This deep insight into information reliability allows the platform to promptly mitigate losses, suspend erroneous automated decisions, and buy valuable time for human intervention, thereby ensuring the continuity of high-tech product production and market competitiveness for enterprises.

[0058] S210. Perform data flow characteristic analysis on the external environment information of material operation; S220. If the data stream characteristics of the material operation external environment information are that the data stream is in a stopped state within a preset time period or the data stream remains unchanged within a preset time period, then the target supplier is determined to be in a failed state, and the first determination result is output. In this embodiment, the platform's internal "heartbeat" monitoring mechanism first sends a lightweight request packet (e.g., an HTTP GET request) to the supplier's data interface periodically (e.g., every 5 seconds), expecting to receive an anticipated confirmation signal within a preset response time (e.g., 200 milliseconds). If no response is received for three consecutive requests, or if the response content does not conform to the preset format, the "heartbeat" monitoring will immediately mark the data source as "no response".

[0059] Secondly, the "Data Flow Behavior Pattern" analysis module continuously records and analyzes the characteristics of data flows from the vendor. For example, it records the average sending interval of data packets over the past 24 hours (e.g., typically millisecond-level updates), the fill rate of key fields (such as inventory quantity, shipment batch) in each data packet, and the normal fluctuation range of these values. If the "heartbeat" monitoring shows no response, and the "Data Flow Behavior Pattern" analysis finds that data transmission has completely stopped (e.g., no data packets have arrived in the past hour), or the values ​​of key fields remain unchanged for a long period, this indicates that the data flow is abnormal.

[0060] At the same time, the external environment information collector also reports that an event has occurred in the supplier's area that could cause its system to crash (e.g., news reports of a strong geomagnetic storm causing a large-scale power outage, or a local government issuing a traffic control notice), and the logistics carrier's system also shows that the freight vehicles dispatched by the supplier have not updated their locations for a long time (e.g., no update of geographical location information for more than 12 hours). Then, the platform will combine all these factors to immediately determine that the core supplier's data source is in a "failed" state.

[0061] S230. If the data flow characteristics of the external environment information of material operation are that there are sporadic information inputs, then a physical mapping consistency assessment is performed on the target supplier, and a first judgment result or a second judgment result is output based on the assessment results. First, it's important to clarify that this embodiment acts like a meticulous "detective" for the industrial internet platform. It no longer relies solely on verbal reports from suppliers but insists on finding evidence from the objective world to verify their authenticity. When a core supplier's information system is paralyzed due to an unforeseen event, and they can only send sporadic messages through temporary and incomplete channels, the platform will not readily believe these messages.

[0062] Specifically, when the platform receives fragmented information from suppliers via non-standard channels, such as "this week's estimated production capacity is down 70%", it immediately activates a dedicated verification mechanism. This mechanism doesn't directly attempt to interpret the precise meaning of the message; instead, it proactively "asks" the external world: has industrial electricity consumption in the supplier's region truly decreased significantly? Does satellite imagery show a genuine reduction in equipment activity at the plant? Have there been large-scale cancellations or delays of raw material orders flowing to the supplier? Only when this objective evidence from independent channels such as power companies, satellite monitoring, and logistics partners logically corroborates the supplier's fragmented report will the platform consider this fragmented information to have preliminary reference value. If this objective evidence contradicts the supplier's report, then even if the information exists, the platform will consider it unreliable and continue to maintain a high level of vigilance regarding related inventory data.

[0063] This approach differs fundamentally from how traditional platforms handle incomplete information. Traditional platforms, upon receiving sporadic information from suppliers via non-standard links, often interpret it as a sign of "partial recovery" of the data source and attempt to adjust the reliability level of inventory data accordingly, even generating "limited" automated decisions based on this incomplete information. For example, when a supplier reports a "70% drop in capacity," a traditional platform might interpret this as "30% capacity remaining," extrapolating inventory and generating allocation suggestions. However, in managing the supply chain of high-value, long-cycle critical chips, supply disruptions have a significant impact, and any decision based on vague or uncertain information can have disastrous consequences. This over-interpretation of "some information is better than none" by traditional platforms often leads to inefficient or even erroneous automated decisions at critical moments, based on flawed judgments of the physical world's true state. This wastes valuable emergency resources and ultimately fails to effectively respond to market changes.

[0064] This embodiment, by insisting on verification through objective phenomena in the physical world, ensures that only information truly reflecting the state of the physical world is used to guide subsequent actions. It recognizes that during periods when supplier information systems are paralyzed and only sporadic information can be transmitted via non-standard links, this information is often delayed, incomplete, and potentially subject to subjective judgment. Therefore, it does not trust any self-reported sporadic information but instead uses rigorous physical mapping verification to minimize the decision-making risks associated with "partially recovered" data. This allows the platform to provide a more robust and realistic judgment when information is extremely uncertain, thus avoiding the acceleration and amplification of erroneous automated decisions in information blind spots. This ensures the continuity and market competitiveness of enterprises in the production of high-tech products, specifically including: S231. If the data stream characteristics of the external environment information of material operation are that there are sporadic information inputs, the sporadic information is identified by keyword body recognition through natural language processing to obtain sporadic key fragment information. S232. Collect physical anchor information corresponding to scattered key fragment information, wherein the physical anchor information includes industrial electricity consumption, satellite remote sensing data, raw material logistics orders and geographical location information; S233. Calculate the physical mapping consistency score between the scattered key fragment information and the physical anchor point information, and obtain the physical mapping consistency score calculation result. S234. Set a threshold for valid information and compare it with the physical mapping consistency score calculation result; S235. If the physical mapping consistency score calculation result is less than the effective information threshold, then the scattered information is regarded as invalid information, and the target supplier is determined to be in a failed state. The first judgment result is output. S236. If the physical mapping consistency score calculation result is greater than the effective information threshold, then the scattered information is regarded as information to be considered, and the second judgment result is output.

[0065] S240. The first or second determination result of the output is used as the determination result of the target supplier status.

[0066] First, it's important to note that during daily operation, the platform continuously sends probe requests to supplier information interfaces periodically via internal "heartbeat" signals, analyzing the transmission patterns of supplier data flows. Simultaneously, the platform operates an "external environment information collector," continuously gathering supplementary information from global news organizations, public social media, and partner logistics carrier systems. When the data flow of a critical processor chip supplier is suddenly interrupted, the platform's data source status verifier immediately activates, cross-referencing the internally detected data flow interruption information with externally collected contextual information such as regional disasters, communication disruptions, and logistical disruptions. Based on this multifaceted and mutually corroborating evidence, the platform determines that the supplier's data source is in a "failed" state and quickly adjusts the "validity weight" of all internal inventory data related to that supplier, reducing it from a completely reliable value to an extremely low level and explicitly marking it as "high-risk." This means that the platform's internal inventory data regarding that chip will no longer be considered completely reliable, and subsequent automated inventory assessments and material allocation decisions will no longer blindly rely on this high-risk data.

[0067] However, within hours of being declared "out of service," the situation began to subtly change. While the communication network in the supplier's region hadn't fully recovered, some backup communication links were temporarily activated. Through these non-standard, low-bandwidth links, the supplier struggled to send non-real-time, aggregated status reports to the platform, such as "This week's estimated production capacity has decreased by 70%" and "Key raw material inventory can support 3 days of service"—macro-level, delayed, and unstructured text messages. The platform's data source status verifier, during continuous monitoring, detected these sporadic, unstructured information streams from the supplier. Although the "heartbeat" monitoring remained unresponsive on the main data interface, and the data flow pattern deviated significantly from normal, the verifier's internal logic interpreted any information received "from the supplier" as a mitigation signal of a "complete data source outage." The external environment information collector also reported vague information that some infrastructure within the region was gradually recovering. Against this backdrop, the platform's data source status verifier's assessment of the supplier's data reliability entered a complex phase. Instead of treating the data source as completely "failed," it fine-tunes its reliability weight from a very low level to a slightly higher intermediate value and labels its status as "partially recovered, but with incomplete information." The internal consideration for this adjustment is that although the data is still unreliable, at least "having information" is better than "having no information at all," and the external environment is also showing signs of recovery.

[0068] In summary, in the context of large manufacturing enterprises managing high-value, long-cycle critical chip supply chains through industrial internet platforms, when the core supplier data source transitions from a complete interruption due to unforeseen events to a partially recovered state through the transmission of sporadic, unstructured, and delayed information via non-standard links, it is crucial to accurately assess the reliability of this "partially recovered" information and its correlation with the physical world's state. This is essential to prevent the platform from over-interpreting incomplete information, which could lead to inefficient automated decision-making based on erroneous judgments and wasted resources. Furthermore, it is vital to ensure that supply chain management can effectively respond to market demands in complex and ever-changing situations.

[0069] Based on this, in this embodiment, when the data source status verifier of the industrial internet platform detects that the data source of the core supplier is in a state of "partial recovery, but incomplete information", the platform will activate a dedicated sporadic information parsing module. This module is responsible for receiving and initially processing sporadic, unstructured information transmitted from the supplier through non-standard, low-bandwidth links.

[0070] Specifically, this sporadic information parsing module can be configured to listen for various non-standard communication protocols and data formats. For example, it can listen for short text messages (SMS or similar protocols) transmitted via satellite phone networks, or key-value pairs sent via emergency data messaging systems (such as MQTT over UDP or custom lightweight protocols). When a text message is received, such as "This week's estimated capacity decrease is 70%", the module uses Natural Language Processing (NLP) technology to identify key entities (such as "capacity", "decline", "70%", "this week") and values, and converts them into structured preliminary information fragments, such as {"indicator": "capacity", "change": "decline", "amplitude": "70%", "time range": "this week"}. For key-value pairs sent via emergency data messaging, the module parses them directly. The purpose of this step is to unify heterogeneous sporadic information into a format that the platform can initially process, but no reliability judgment is performed at this time.

[0071] Furthermore, the platform initiates a physical mapping verifier, which actively searches for independent verification points (i.e., "physical anchor points") that may exist in the physical world based on the parsed fragmented information from the external environment information collector. The specific implementation is as follows: 1) Industrial Electricity Consumption Monitoring: If sporadic information involves "capacity decline" or "production stagnation," the physical mapping validator immediately requests industrial electricity consumption data for the supplier's region from the external environmental information collector. The external environmental information collector connects to the local power company's data interface (e.g., real-time data queries based on a RESTful API) to obtain industrial electricity consumption curves for the region over the past 24 hours or longer. The validator calculates the percentage deviation of the current period's industrial electricity consumption from the normal level before the disaster. For example, if sporadic information reports "capacity decline of 70%" while industrial electricity consumption has only declined by 20%, there is a significant discrepancy between the two.

[0072] 2) Satellite Remote Sensing Data Analysis: For information concerning plant activities or equipment operation, the verifier requests satellite remote sensing data. The external environment information collector can integrate APIs from commercial satellite imagery services (such as Maxar Technologies or Planet Labs) to obtain the latest high-resolution satellite imagery or thermal imaging data of the supplier's plant. The physical mapping verifier analyzes the frequency of vehicle entry and exit in the images, the thermal signal intensity of equipment (e.g., through pixel brightness or infrared spectral analysis), and compares it with baseline data before the disaster. For example, if sporadic information suggests a significant reduction in production activity, but satellite thermal imaging shows that the thermal signal intensity of the plant's main production equipment is not significantly different from normal operating conditions, it indicates that the information does not match physical reality.

[0073] 3) Raw Material Logistics Order Tracking: If sporadic information relates to "critical raw material inventory" or "production support days," the validator will query the status of critical raw material logistics orders flowing to that supplier via an external environment information collector. The external environment information collector connects to the upstream raw material supplier's logistics system (e.g., via EDI or API) to obtain recent cancellation, delay, or modification records of raw material orders shipped to that core supplier. The physical mapping validator will then calculate the proportion of large-scale order cancellations or delays. For example, if a supplier reports "critical raw material inventory can support 3 days," but there are no large-scale cancellations or delays of raw material orders flowing to that supplier, and new orders are even being shipped normally, this may indicate a discrepancy in the supplier's reporting.

[0074] 4) Geographic Location Information Comparison: If sporadic information involves "shipped" or "transported," the verifier compares the geographic location information of the materials in transit provided by the logistics carrier's system. The external environment information collector continuously obtains real-time GPS location data for all materials in transit from the cooperating logistics carrier's system. The physical mapping verifier checks whether the "shipped" or "transported" status reported by the supplier matches the actual material movement trajectory. For example, if the supplier reports "shipped," but the logistics system shows that the relevant batch of materials has been staying near the supplier's warehouse for an extended period, there is an inconsistency.

[0075] Furthermore, in the physical mapping consistency assessment and reliability weight adjustment, the physical mapping verifier performs a logical consistency assessment between the parsed fragments of information and all collected physical anchor information. The specific implementation method is as follows: 1) Consistency Score Calculation: For each fragment of information, the validator calculates a "Physical Mapping Consistency Score" (PMC) based on its consistency with the information of each physical anchor point. The formula for calculating PMC is as follows:

[0076] in, Indicates the first Consistency index between physical anchor points and scattered information (e.g., 0 indicates complete inconsistency, 0.5 indicates partial consistency, and 1 indicates complete consistency). Indicates the first The importance weight of each physical anchor point (for example, the weight of industrial electricity consumption on the impact of production capacity may be higher than that of satellite remote sensing data).

[0077] For example, if a supplier reports a 70% decrease in production capacity, while industrial electricity consumption has decreased by 65% ​​( Satellite thermal imaging showed a significant reduction in activity within the factory area. Raw material orders experienced minor delays but no large-scale cancellations. ), then PMC = (0.5 0.9 + 0.3 0.8 + 0.2 0.4) / (0.5 + 0.3 + 0.2) = (0.45 +0.24 + 0.08) / 1 = 0.77.

[0078] 2) Valid Information Determination: The verifier sets a "valid information threshold" (e.g., PMC > 0.7). Only when the PMC is higher than this threshold will the sporadic information be preliminarily deemed to have "valid information" status. Otherwise, even if there is information, it is considered "informative but unreliable".

[0079] S300: Based on the target supplier status determination result, dynamically adjust the reliability weight, and assign a value to the validity weight of the inventory data according to the dynamic weight adjustment result, so as to realize dynamic management of supply chain inventory.

[0080] S310. Based on the target supplier status determination results, dynamically adjust the reliability weight; S320. If the target supplier status determination result is the first determination result, then set the inventory data validity weight corresponding to the target supplier to the first preset value, and obtain the adjustment result of the first inventory data validity weight. In this embodiment, based on this determination, the platform dynamically adjusts the "validity weight" of all internal inventory data related to that supplier. This weight can be a value between 0 and 1, initially set to 1 (fully reliable). When the data source is determined to be "failed," the weight will quickly drop to 0.1 or lower, and be explicitly marked as "high-risk." For example, if the inventory data of a critical chip was originally considered 100% reliable, once its supplier data source is determined to be failed, the reliability weight of the chip's inventory data will immediately be adjusted to 10%, and the system will label it as "high-risk" in red, rather than simply treating it as a "temporary communication failure" awaiting recovery. This multi-dimensional cross-validation mechanism effectively avoids the platform misjudging a regional disaster causing the upstream critical chip supplier system to be paralyzed as a temporary communication failure when facing a global layout, thus providing accurate and timely basis for subsequent decision-making adjustments.

[0081] S330. If the target supplier status determination result is the second determination result, then based on the physical mapping consistency score calculation result, the reliability weight of the inventory data validity weight corresponding to the target supplier is dynamically adjusted to obtain the updated inventory data validity weight value, and the adjustment result of the second inventory data validity weight is obtained. The updated inventory data validity weight value is greater than the first preset value. In this embodiment, the reliability weight is dynamically adjusted. Based on the PMC score and the effective information determination result, the platform dynamically adjusts the "effectiveness weight" of all internal inventory data related to the supplier. If the PMC is below the threshold, even if there is sporadic information, the weight remains at an extremely low level (e.g., 0.1) and continues to be marked as "high risk". If the PMC is above the threshold, the weight can be fine-tuned according to the PMC value. For example, the new effectiveness weight = original extremely low weight + (PMC - effective information threshold) (intermediate value - original extremely low weight) / (1 - effective information threshold), the specific expression of which is as follows:

[0082] In the above formula, This represents the validity weight value of the updated inventory data. This represents the first preset value. Indicates the physical mapping consistency score. Indicates the threshold of valid information. This represents the midpoint between the effective information threshold and the first preset value.

[0083] For example, if the original weight is 0.1, the median value is 0.3, the effective information threshold is 0.7, and the PMC is 0.77, then the new weight = 0.1 + (0.77 - 0.7) (0.3 - 0.1) / (1 - 0.7) = 0.1 + 0.07 0.2 / 0.3 = 0.1 + 0.046 = 0.146. Simultaneously, the platform will update its status label, for example, changing it from "high risk" to "risk mitigated, but still requires manual review".

[0084] Finally, the decision-making logic is adjusted and manual intervention is triggered. When the "validity weight" of internal inventory data is adjusted based on the physical mapping consistency assessment results, the platform will adjust its automated decision-making logic accordingly. The implementation method is as follows: 1) Automated decision suspension or restriction: If the validity weight remains at an extremely low level (e.g., 0.1), the platform will continue to suspend all automated allocations and production plan adjustments that rely on the supplier's data.

[0085] 2) Limited Automated Suggestions: If the validity weight is slightly increased (e.g., 0.146), the platform can generate "limited" automated suggestions, but these suggestions will not be executed automatically. For example, the platform can suggest "purchasing a small number of secondary critical components from alternative suppliers within the next three days" based on sporadic information and capacity changes calculated from physical anchors, but this suggestion will come with a clear risk warning and require mandatory manual approval.

[0086] 3) Mandatory Manual Review and Physical Inventory: Regardless of how validity weights are adjusted, the platform will forcibly trigger a manual review process as long as the data source is in a "partially recovered" state. For example, an emergency alert will be sent to the purchasing manager, supply chain director, and warehouse center manager, requiring them to conduct a physical inventory check of the relevant materials. Based on the physical mapping consistency assessment report provided by the platform, a manual judgment will be made as to whether to initiate procurement from alternative suppliers or adjust the production plan. The platform will provide an interactive interface displaying miscellaneous information, all physical anchor data, and their consistency scores to assist in manual decision-making.

[0087] S340. The adjustment result of the first inventory data validity weight or the adjustment result of the second inventory data validity weight is used as the final inventory data validity weight adjustment result to realize dynamic management of supply chain inventory.

[0088] In summary, traditional systems, when data flow is interrupted, often only report "connection lost" or "data missing," then passively use the last acquired information or wait for data recovery. This approach is particularly vulnerable to sudden, external, supply chain disruptions that impact upstream data sources. For example, when a globally positioned company experiences a complete information system paralysis due to a regional disaster affecting the supply of high-value, long-cycle critical chips, traditional platforms may misjudge it as a temporary communication failure, continuing to rely on outdated "adequate inventory" assumptions and accelerating the generation of erroneous cross-regional material transfer instructions, ultimately leading to production shutdowns.

[0089] This invention introduces a proactive "contextual awareness" capability and a multi-dimensional "information fusion" mechanism. It goes beyond simply focusing on the existence of data, delving deeper into the "vitality" and "authenticity" of the data. By integrating "heartbeat" monitoring, data flow behavior pattern analysis, and proactively monitoring tracking data from global news media, social media, and logistics carriers, the platform can seek clues from a broader external environment and conduct comprehensive cross-validation of the health status of data sources. This deep insight into the "life state" of data enables the platform to shift from passive waiting to proactive risk identification. It can distinguish between two fundamentally different data interruption scenarios: "network jitter" and "upstream systemic collapse," treating "no data" as a clear risk signal, rather than merely a technical malfunction.

[0090] This approach, which incorporates extensive information from the "external world" into supply chain decision-making, breaks the conventional thinking that "inventory data can only come from direct reports from suppliers." It enables the platform to build a more comprehensive "risk profile" when core data sources fail, providing a unique "balance point" through cross-verification of multiple supporting information sources. This balance point lies in reducing decision-making risk by "enhancing the perception of uncertainty" when accurate "real-time data" is unavailable. It doesn't attempt to directly repair broken data flows, but rather buffers the impact of core data failure by building an "external information firewall." This emphasis on the diversity of information sources, and the strategy of utilizing "unofficial" information for decision-making during crises, allows the platform to avoid blind decisions when facing fundamental disruptions in the upstream supply chain. Instead, it enables judgments based on richer and more verifiable information, transforming localized data problems into controllable risks and effectively preventing global production shutdowns and huge economic losses.

[0091] Therefore, the embodiments of the present invention can significantly improve the risk identification and response capabilities of industrial internet platforms when faced with external data source interruptions. Through multi-dimensional cross-validation of information, it ensures that the platform can accurately identify the failure state of the data source, rather than misjudging it as a temporary communication failure. This enables the platform to dynamically adjust the reliability level of internal inventory data, promptly suspend erroneous automated decisions, and force manual intervention, thereby effectively avoiding ineffective allocations, production stoppages, and huge economic losses caused by information asymmetry, and ensuring the resilience and production continuity of the enterprise supply chain.

[0092] Reference Figure 2 A dynamic supply chain inventory management system based on an industrial internet platform includes: The first module 201 is used to acquire information on natural emergencies and freight status related to supply chain operations, and to perform uncertainty quantification assessment and correction on freight status information to construct external environment information for material operation. The second module 202 is used to perform data flow feature analysis and physical mapping consistency assessment on the external environment information of material operation to obtain the target supplier status determination result. The third module 203 is used to dynamically adjust the reliability weight based on the target supplier status determination result, and assign a value to the validity weight of the inventory data based on the dynamic weight adjustment result, so as to realize dynamic management of supply chain inventory.

[0093] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0094] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for dynamic supply chain inventory management based on an industrial internet platform, characterized in that, Includes the following steps: Acquire information on natural emergencies and freight status related to supply chain operations, and conduct uncertainty quantification assessment and correction of freight status information to construct information on the external environment of material movement; Data flow feature analysis and physical mapping consistency assessment are performed on the external environment information of material operation to obtain the target supplier status determination result; Based on the target supplier status assessment, the reliability weight is dynamically adjusted, and the validity weight of the inventory data is assigned according to the dynamic adjustment result, thereby realizing dynamic management of supply chain inventory.

2. The supply chain inventory dynamic management method based on an industrial internet platform according to claim 1, characterized in that, The step of acquiring information on natural emergencies and freight status related to supply chain operations, and performing uncertainty quantification assessment and correction on freight status information to construct information on the external environment of material movement, specifically includes: Acquire information on natural emergencies and freight status related to supply chain operations. Natural emergencies include natural disasters, regional power outages, communication network failures, or traffic control. Freight status information includes material transport speed and estimated arrival time. Uncertainty quantification and correction are performed on freight status information to obtain corrected freight status information; By cross-integrating natural emergencies with the corrected freight status information, we can construct external environment information for material operation.

3. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 2, characterized in that, The step of performing uncertainty quantification assessment and correction on freight status information to obtain corrected freight status information specifically includes: The multi-source physical world information of the carrier corresponding to the freight status information is obtained. The multi-source physical world information includes traffic flow data, road monitoring images, traffic flow sensing speed data and comparison frequency. The comparison frequency represents the comparison frequency between the material transportation speed in the freight status information and the traffic flow sensing speed data in the multi-source physical world information within a preset time range. The deviation between the material transport speed in the freight status information and the multi-source physical world information is quantified and tracked to obtain the initial deviation value and the deviation duration. Uncertainty scores are calculated based on the initial deviation value and the duration of the deviation. The estimated arrival time in the freight status information is then corrected based on the results of the uncertainty score calculation to obtain the corrected freight status information.

4. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 3, characterized in that, The specific expression for calculating the uncertainty score is as follows: In the above formula, Represents the uncertainty score. This represents the initial deviation value. Indicates the duration of the deviation. Indicates the preset time unit. This indicates a preset, persistent impact factor.

5. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 1, characterized in that, The step of performing data flow feature analysis and physical mapping consistency assessment on the external environment information of material operation to obtain the target supplier status determination result specifically includes: Perform data flow characteristic analysis on information about the external environment of material operation; If the data stream characteristics of the material operation external environment information are that the data stream is in a stopped state within a preset time period or the data stream remains unchanged within a preset time period, then the target supplier is determined to be in a failed state, and the first determination result is output. If the data flow characteristics of the material operation external environment information are that there are sporadic information inputs, then a physical mapping consistency assessment is performed on the target supplier, and a first judgment result or a second judgment result is output based on the assessment results. The first or second determination result of the output is used as the determination result of the target supplier status.

6. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 5, characterized in that, The step of performing a physical mapping consistency assessment on the target supplier and outputting a first or second judgment result based on the assessment result if the data stream characteristics of the material's external environment information show sporadic information input includes: If the data stream characteristics of the material operation external environment information are that there are sporadic information inputs, the sporadic information can be identified by keyword body recognition through natural language processing to obtain sporadic key fragment information. Collect physical anchor information corresponding to scattered key fragment information, including industrial electricity consumption, satellite remote sensing data, raw material logistics orders, and geographical location information; A physical mapping consistency score is calculated between scattered key fragment information and physical anchor point information to obtain the physical mapping consistency score calculation result; Set a threshold for valid information and compare it with the physical mapping consistency score calculation results; If the physical mapping consistency score is less than the effective information threshold, the scattered information is considered invalid information, the target supplier is determined to be in a failed state, and the first determination result is output. If the physical mapping consistency score is greater than the effective information threshold, then the scattered information is considered as information to be considered, and the second judgment result is output.

7. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 6, characterized in that, The specific expression for calculating the physical mapping consistency score is as follows: In the above formula, This represents the result of the physical mapping consistency score calculation. Indicates the first Consistency index between physical anchor points and scattered information Indicates the first The importance weight of each physical anchor point.

8. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 6, characterized in that, The step of dynamically adjusting reliability weights based on the target supplier status determination results, and assigning values ​​to inventory data validity weights based on the dynamic weight adjustment results to achieve dynamic supply chain inventory management, specifically includes: Based on the target supplier status assessment results, the reliability weight is dynamically adjusted; If the target supplier status determination result is the first determination result, then the inventory data validity weight corresponding to the target supplier is set to the first preset value, and the adjustment result of the first inventory data validity weight is obtained; If the target supplier status determination result is the second determination result, then the reliability weight is dynamically adjusted based on the physical mapping consistency score calculation result, and the inventory data validity weight corresponding to the target supplier is obtained by updating the inventory data validity weight value, and the second inventory data validity weight adjustment result is obtained. The updated inventory data validity weight value is greater than the first preset value. The final inventory data validity weight adjustment result is obtained by adjusting either the first or second inventory data validity weight based on the output, thus achieving dynamic inventory management in the supply chain.

9. The method for dynamic supply chain inventory management based on an industrial internet platform according to claim 8, characterized in that, The specific expression for the dynamic adjustment of the reliability weight is as follows: In the above formula, This represents the validity weight value of the updated inventory data. This represents the first preset value. Indicates the physical mapping consistency score. Indicates the threshold of valid information. This represents the midpoint between the effective information threshold and the first preset value.

10. A dynamic supply chain inventory management system based on an industrial internet platform, characterized in that, Includes the following modules: The first module is used to acquire information on natural emergencies and freight status related to supply chain operations, and to perform uncertainty quantification assessment and correction on freight status information to construct information on the external environment of material operation. The second module is used to perform data flow feature analysis and physical mapping consistency assessment on the external environment information of material operation to obtain the target supplier status determination result. The third module is used to dynamically adjust the reliability weight based on the target supplier status determination result, and assign a value to the validity weight of the inventory data based on the dynamic weight adjustment result, so as to realize dynamic management of supply chain inventory.

Images