CN-120564990-B - Full-link supply chain data management method and system based on digital twinning

CN120564990BCN 120564990 BCN120564990 BCN 120564990BCN-120564990-B

Abstract

The invention provides a full-link supply chain data management method and system based on digital twinning, wherein the method comprises the steps of collecting real-time data from a plurality of supply chain data sources, inputting the data into a digital twinning model, storing the data in a distributed database, carrying out local data preprocessing based on edge computing nodes on the digital twinning model, designing a comprehensive objective function, dynamically adjusting the weight of each optimization objective according to real-time environment change and task requirements, ensuring that each task reasonably distributes resources according to the priority, designing a three-dimensional visual system, displaying the optimization task and the resource state in a three-dimensional space, and displaying the link states, the optimization effects and the potential problems of the supply chain in real time through a three-dimensional visual interface. The invention adopts the leading edge technologies such as digital twin technology, edge calculation, self-adaptive optimization algorithm, cloud edge cooperation and the like, and solves a plurality of pain points in the traditional medical supply chain management.

Inventors

WANG YINHUI
LI MINSHENG
TANG WENJUAN
YANG WEN

Assignees

广州基医云计算有限公司

Dates

Publication Date: 20260512
Application Date: 20250418

Claims (8)

1. A method for managing full link supply chain data based on digital twinning, the method comprising the steps of: Real-time data acquisition is carried out from a plurality of supply chain data sources, a digital twin model is built according to the acquired data sources, wherein each data source transmits data to a cloud through an edge computing node, and the data are input into the digital twin model and stored in a distributed database; local data preprocessing based on an edge computing node is carried out on the digital twin model, encryption processing based on symmetric encryption is carried out on the preprocessed data, and after the encryption processing, the edge computing node transmits the encrypted data to a cloud; the method comprises the steps of collecting preprocessed data sources, carrying out weighted aggregation on data from different sources by an edge computing node to obtain a single data set, wherein when the accuracy of one data source is higher in the single data set, the weight value of the single data set is correspondingly increased, so that the fused data is ensured to be more accurate; according to the real-time data and the calculation result of the edge calculation node, designing a comprehensive objective function, wherein the comprehensive objective function is used for measuring the scheduling effect, dynamically adjusting the weight of each optimization target according to the real-time environment change and the task demand, and ensuring that each task reasonably distributes resources according to the priority of each task; According to the comprehensive objective function and the priority, designing a three-dimensional visual system to display the optimized task and the resource state in a three-dimensional space, and displaying the state, the optimized effect and the potential problem of each link of a supply chain in real time through a three-dimensional visual interface, wherein three coordinate axes of the three-dimensional visual system respectively represent the task type, the resource consumption and the timeliness of task completion; the method comprises the steps of determining a target loss based on the comprehensive objective function, wherein the comprehensive objective function comprises a plurality of objective functions, total cost and delay loss of task execution, and the objective function represents efficiency of task execution; The dynamically adjusting the weight of each optimization target according to the real-time environment change and the task demand comprises the following steps: the weight environment evaluation function of each optimization target evaluates the priority of different targets according to real-time feedback, and adjusts the weight of each optimization target according to the priority of different targets, wherein the weight environment evaluation function is an environment change value evaluated according to real-time data; Meanwhile, after the target loss is determined, a delay penalty term is also required to be calculated, and the scheduling sequence of the tasks is optimized by introducing the delay penalty term, so that resource waste or efficiency reduction caused by scheduling delay is reduced; The delay penalty term penalizes tasks with longer delay, so that the system is forced to process urgent or delay tasks preferentially, and optimal allocation of resources is ensured.
2. The digital twinning-based full link supply chain data management method of claim 1, wherein the real-time data includes order information, inventory status, transportation track, and environmental monitoring; the construction of the digital twin model is based on a data source of a supply chain, the data source forms a state vector of each link, and a mapping function is used for constructing the state vector of each link.
3. The method for managing data of a full link supply chain based on digital twinning according to claim 2, wherein each data source transmits data to a cloud end through an edge computing node, and inputs the data into a digital twinning model, and stores the data in a distributed database, comprising: Each data source transmits data to the cloud through the edge computing node, when the data source is delayed in transmission, the digital twin model is updated in real time by using a correction function of the digital twin model state by the data source, and the digital twin model state is ensured to be consistent with the operation state of the real world at each moment; if data loss, abnormal fluctuation or noise interference occurs in the actual data acquisition process, executing: triggering an early warning mechanism, and automatically adjusting the state of the digital twin model to recalibrate the digital twin model.
4. The digital twinning-based full link supply chain data management method of claim 1, wherein the local data preprocessing comprises: the data noise removing process is adjusted based on the deviation between the abnormal value of the current data and the distribution of the historical data; and (3) filling missing data points by adopting a time sequence interpolation method under the condition of data missing based on a data source of the digital twin model, so as to keep the continuity and consistency of the data.
5. The digital twinning-based full link supply chain data management method according to claim 1, wherein said ensuring that each task allocates resources rationally according to its priority comprises: in the task execution process, feedback data of the task execution condition are collected in real time, dynamic adjustment is carried out according to the feedback, meanwhile, the comprehensive objective function and the resource allocation are continuously optimized through an error function, so that continuous optimization of a scheduling target is achieved, the error function is used for measuring the gap between a current scheduling result and an expected target, in the optimization process, errors are minimized, and the weight of each optimization target, the priority of different targets and the resource allocation are continuously adjusted, so that each scheduling is closer to the expected target, and the effect of self optimization is achieved.
6. The digital twinning-based full link supply chain data management method according to claim 5, wherein the coordinate system of the three-dimensional space is Wherein, the The axis represents the task type and, The axis represents the consumption of resources and, The axis represents timeliness of task completion; The states of the tasks are represented by color codes, the tasks with longer delay are displayed in red, the emergency tasks are displayed in yellow, and the normal tasks are displayed in green; meanwhile, a dynamic priority model is designed according to a coordinate system of a three-dimensional space, and the priority is adjusted by evaluating the space coordinates and the resource occupation condition of a task in real time; wherein the dynamic priority model is task-based Basic priority, task of (c) Coordinates in three-dimensional space, and tasks Coordinate determination in three-dimensional space; the dynamic priority model can dynamically adjust the priority of the task according to the position of the task in the three-dimensional space, the priority of the urgent task is higher, and the priority is highest when the spatial distance between the urgent task and other tasks is closest.
7. The method for managing data of a full link supply chain based on digital twinning according to claim 6, wherein the decision support provided by the three-dimensional visualization system is based on optimized task scheduling data, dynamic decision adjustment is performed in combination with real-time feedback, a proposed adjustment scheme can be provided based on a current task state and an optimization target, and the decisions of scheduling optimization and resource allocation can be performed in real time through the cooperation of the three-dimensional visualization system and the decision support.
8. A full link supply chain data management system based on digital twinning according to claim 1, wherein the system comprises: The system comprises a supply chain data source acquisition unit, a distributed database and a distributed data storage unit, wherein the supply chain data source acquisition unit is used for acquiring real-time data from a plurality of supply chain data sources, and constructing a digital twin model according to the acquired data sources, wherein each data source transmits data to a cloud through an edge computing node, and inputs the data into the digital twin model and stores the data in the distributed database; The supply chain data source analysis unit is used for carrying out local data preprocessing based on the edge computing node on the digital twin model, carrying out encryption processing based on symmetric encryption on the preprocessed data, and transmitting the encrypted data to the cloud end by the edge computing node after the encryption processing; The resource allocation unit is used for designing a comprehensive objective function according to the real-time data and the calculation result of the edge calculation node, wherein the comprehensive objective function is used for measuring the scheduling effect, dynamically adjusting the weight of each optimization target according to the real-time environment change and the task demand, and ensuring that each task reasonably allocates resources according to the priority of each task; The visual display unit is used for designing a three-dimensional visual system to display the optimized task and the resource state in a three-dimensional space, and displaying the state, the optimized effect and the potential problem of each link of the supply chain in real time through a three-dimensional visual interface, wherein three coordinate axes of the three-dimensional visual system respectively represent the task type, the resource consumption and the timeliness of task completion.

Description

Full-link supply chain data management method and system based on digital twinning Technical Field The invention belongs to the field of digital twinning, and particularly relates to a full-link supply chain data management method and system based on digital twinning. Background In modern medical supply chain management, along with rapid growth of medical demands and increasing complexity of supply chain management, ensuring timely, accurate and safe supply of medicines, medical instruments and other medical supplies is a core problem to be solved urgently. Conventional medical supply chains typically involve multiple links, including production, warehousing, transportation, distribution, inventory management, etc., that require rapid and efficient information flow and logistics support therebetween. However, existing medical supply chain management techniques often face several challenges that limit their efficiency and flexibility. First, existing supply chain management systems have the problem of information islanding in terms of data processing and scheduling optimization. Many medical institutions and suppliers use independent systems such as ERP (enterprise resource planning), SCM (supply chain management), etc., and there is insufficient information sharing between these systems. The information of each link is isolated in different systems, so that the data exchange is unsmooth, the decision is delayed, and even at the key moment, the situation of excessive or shortage of stock is generated due to the lack of accurate data support. The problem of such information islands severely affects the overall performance of the supply chain, particularly in the medical industry where timely delivery and accurate inventory management of materials is directly related to patient treatment and life safety. Secondly, with the development of internet and internet of things technologies, a great amount of real-time data (such as inventory status, transportation track, environmental condition, order information, etc.) in a medical supply chain gradually becomes an important component of management. However, conventional centralized computing architectures (e.g., data centers and servers) do not take full advantage of edge computing technology to increase the processing power of real-time data. Under the traditional architecture, data needs to be processed and stored through a central server, and delay is often generated in the process, and particularly when a large amount of real-time dynamic data is faced, efficient instant decision making is difficult to realize. This is particularly true in the face of highly aged, high risk materials such as pharmaceuticals, medical devices, etc. Because the response speed of the traditional system is low, effective adjustment cannot be timely performed on the supply chain under emergency conditions such as fluctuation of demand, transportation delay or insufficient inventory, and the stability and emergency response capability of the whole supply chain are affected. Furthermore, as the management of the medical supply chain becomes increasingly complex and diversified, how to realize the collaborative optimization of each link becomes a long-felt problem in the industry. The traditional supply chain optimization method is mostly based on a static optimization model, but lacks the capability of flexible and dynamic adjustment. For example, when an emergency (such as insufficient warehouse, delayed transportation, fluctuation of demand, etc.) occurs in a certain link in the supply chain, the conventional optimization method cannot quickly respond to the dynamic changes, and lacks the capability of intelligent self-adaptive adjustment. In particular, medicines, vaccines and other materials with strict requirements on humiture and transportation timeliness, the conventional supply chain management system often cannot monitor and optimize in real time, so that risks of stock backlog or overdue loss can occur. In addition, the conventional visual interface is limited to two-dimensional or static display, and is difficult to comprehensively and intuitively reflect the multi-dimensional and dynamic changes of the supply chain. The existing visual interface can only display the data of a certain link, and a manager can hardly acquire the global information of each link in a short time, so that information asymmetry and decision errors are inevitably generated when scheduling decisions are made. Therefore, although the prior art has advanced in some aspects, they still cannot effectively solve a series of key problems of real-time, flexibility, synergy, and visualization in the medical supply chain management, and an innovative solution for integrating advanced technology and comprehensively improving the efficiency and response speed of the supply chain is needed. Disclosure of Invention The invention aims to provide a full-link supply chain data management method and system based on di