Medical digital supply chain risk analysis method, system, equipment and product

Abstract

The invention belongs to the technical field of supply chain risk control, and aims to provide a medical digital supply chain risk analysis method, system, equipment and product. According to the method, the supply chain graph and the time sequence prediction model are utilized to dynamically update the node attribute information and perform subsequent participant node risk score calculation, so that the dynamic change in the specified supply chain structure can be comprehensively captured, the sensitivity of risk early warning is enhanced, and the risk early warning efficiency is improved. The requirements of real-time monitoring and quick response of the whole process of the supply chain can be met; besides, setting of a graph convolutional network is introduced into risk propagation modeling, so that a potential risk path can be effectively identified, the problem of difficulty in risk identification in a complex scene in the prior art is solved, and efficient circulation and safe supply of medical supplies in each link of a supply chain are ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of supply chain risk control, and specifically relates to a medical digital supply chain risk analysis method, system, equipment and product. Background Art

[0002] The medical digital supply chain is an important part of the modern medical industry to achieve efficient management and optimize resource allocation. It is widely used in the fields of medical material procurement, production, logistics, inventory management and distribution. At present, the digital construction of the medical supply chain mainly relies on the Internet of Things, big data analysis and artificial intelligence technology. The real-time collection of material status, location and environmental information through Internet of Things devices; combined with big data analysis technology, to explore potential patterns and optimization opportunities in the supply chain; using artificial intelligence models to predict demand changes, optimize inventory allocation, and perform anomaly detection and risk warning. These technologies provide effective digital support for the supply chain management of the medical industry, significantly improving management efficiency and decision-making capabilities.

[0003] However, in the process of using the prior art, the inventors found that the prior art has at least the following problems:

[0004] Existing supply chain risk monitoring is usually based on rule models or traditional machine learning methods. These methods have limited adaptability to complex and changeable supply chain scenarios and cannot effectively capture potential risk patterns. In the event of emergencies (such as epidemics, natural disasters) or drastic changes in the supply chain environment, the risk warning sensitivity is insufficient, which often leads to delayed risk response or false warnings. Summary of the invention

[0005] The present invention aims to solve the above-mentioned technical problems at least to a certain extent. The present invention provides a medical digital supply chain risk analysis method, system, equipment and product.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a medical digital supply chain risk analysis method, comprising:

[0008] Acquire initial attribute information of each participant node in the specified supply chain, and construct a supply chain graph with each participant node in the specified supply chain as an entity and the connection relationship between each participant node in the specified supply chain as an edge;

[0009] Collect multi-source heterogeneous data of the specified supply chain in real time, and perform data fusion processing on the multi-source heterogeneous data to obtain fused data;

[0010] Acquire multiple fused data of the specified supply chain at k historical time steps before the current moment to obtain a fused data set, and input the fused data set into a preset time series prediction model to obtain a time series prediction result;

[0011] According to the time series prediction result, the initial attribute information of each participant node in the specified supply chain is updated to obtain updated attribute information of each participant node;

[0012] Based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network is used to obtain a risk score of each participant node in the supply chain graph;

[0013] From each participant node, the participant nodes whose risk scores are greater than a preset risk threshold are screened out, and based on the participant nodes whose risk scores are greater than the risk threshold, a high-risk participant list is constructed.

[0014] In one possible design, the supply chain graph is G = (V, E), where V = {v 1 ,v 2 ,...,v n},v 1 ,v 2 ,...,v n represents the nodes of each participant in the specified supply chain, E = {e ij |v i ,v j ∈V}, indicating the connection relationship between the nodes of each participant.

[0015] In a possible design, the multi-source heterogeneous data includes logistics data D IoT , purchasing data D ERP and external environment data D ext , the fused data is:

[0016] D fused =w 1 D IoT +w 2 D ERP +w 3 D ext ;

[0017] In the formula, w 1 is the weight of logistics data, w 2 is the weight of the purchase data, w 3 is the weight of external environment data.

[0018] In one possible design, the fused dataset is represented as: t ={D fused,t-k , D fused,t-k+1, ..., D fused,t}; where k is the time window of the fused data set, D fused,t-k , D fused,t-k+1 , ..., D fused,t is the fused data at time tk, t-k+1, ... and t in the fused data set;

[0019] Using the Transformer-based time series model, the time series prediction results are: In the formula, X t is the fused data set, and θ is the preset model parameter.

[0020] In a possible design, based on the supply chain graph and updated attribute information of each participant node, a graph convolutional network is used to obtain a risk score of each participant node in the supply chain graph, including:

[0021] Obtaining an adjacency matrix matching the supply chain graph;

[0022] The adjacency matrix and the updated attribute information of each participant node are used as input data of a graph convolutional network, so as to update the risk characteristics of the participant nodes in the supply chain graph through the graph convolutional network, thereby obtaining the node characteristics of each participant node in the supply chain graph;

[0023] According to the node characteristics of each participant's node, the risk score of each participant's node is obtained through a preset linear mapping function.

[0024] In a possible design, the formula for updating the risk characteristics of the participant nodes in the supply chain graph through the graph convolutional network is as follows:

[0025] H (l+1) =σ(AH (l) W (l) );

[0026] Where A is the adjacency matrix matching the supply chain graph, H (l) is the node feature output by the lth network layer in the graph convolutional network, l is a natural number greater than or equal to 0, H (l+1) is the node feature output by the l+1th network layer in the graph convolutional network, W (l) is the weight matrix of the lth network layer in the graph convolutional network, and σ is the preset activation function.

[0027] In a second aspect, the present invention provides a medical digital supply chain risk analysis system, comprising:

[0028] A graph construction module is used to obtain initial attribute information of each participant node in a specified supply chain, and construct a supply chain graph by taking each participant node in the specified supply chain as an entity and the connection relationship between each participant node in the specified supply chain as an edge;

[0029] An attribute updating module is communicatively connected with the graph building module, and is used to collect multi-source heterogeneous data of the specified supply chain in real time, and perform data fusion processing on the multi-source heterogeneous data to obtain fused data; is used to obtain multiple fused data of the specified supply chain at k historical time steps before the current moment to obtain a fused data set, and input the fused data set into a preset time series prediction model to obtain a time series prediction result; and is also used to update the initial attribute information of each participant node in the specified supply chain according to the time series prediction result to obtain the updated attribute information of each participant node;

[0030] A risk score acquisition module, which is in communication with the attribute update module and is used to obtain the risk score of each participant node in the supply chain graph using a graph convolutional network based on the supply chain graph and the updated attribute information of each participant node;

[0031] The node screening module is communicatively connected with the risk score acquisition module, and includes screening out the participant nodes whose risk scores are greater than a preset risk threshold from the participant nodes, and constructing a high-risk participant list based on the participant nodes whose risk scores are greater than the risk threshold.

[0032] In a third aspect, the present invention provides an electronic device, comprising:

[0033] a memory for storing computer program instructions; and,

[0034] A processor is used to execute the computer program instructions to complete the operation of a medical digital supply chain risk analysis method as described in any one of the above.

[0035] In a fourth aspect, the present invention provides a computer program product, including a computer program or instructions, which, when executed by a computer, implements a medical digital supply chain risk analysis method as described in any one of the above.

[0036] In a fifth aspect, the present invention provides a computer-readable storage product having instructions stored thereon. When the instructions are run on a computer, a medical digital supply chain risk analysis method as described in any one of the above is executed.

[0037] The beneficial effects of the present invention are:

[0038] The present invention discloses a medical digital supply chain risk analysis method, system, equipment and product, which can realize risk monitoring and early warning of the supply chain with high real-time performance and accuracy. Specifically, during the implementation of the present invention, the initial attribute information of each participant node in the specified supply chain is obtained in advance, and the node of each participant in the specified supply chain is taken as an entity, and the connection relationship between the nodes of each participant in the specified supply chain is taken as an edge to construct a supply chain graph; multi-source heterogeneous data of the specified supply chain is collected in real time, and the multi-source heterogeneous data is subjected to data fusion processing to obtain fused data, and then multiple fused data of the specified supply chain at k historical time steps before the current moment are obtained to obtain a fused data set, and the fused data set is input into a preset time series prediction model to obtain a time series prediction result; then, according to the time series prediction result, the initial attribute information of each participant node in the specified supply chain is updated to obtain the updated attribute information of each participant node; then, based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network is used to obtain the risk score of each participant node in the supply chain graph; finally, from each participant node, a participant node with a risk score greater than a preset risk threshold is screened, and a high-risk participant list is constructed based on the participant node with a risk score greater than the risk threshold. In the present invention, the supply chain graph and time series prediction model are used to dynamically update the node attribute information and perform subsequent calculations on the risk scores of the participating nodes, which can help to comprehensively capture the dynamic changes in the specified supply chain structure, enhance the sensitivity of risk warning, and meet the needs of real-time monitoring and rapid response to the entire supply chain process; in addition, the introduction of the graph convolutional network setting in risk propagation modeling can effectively identify potential risk paths, solve the problem of difficulty in risk identification in complex scenarios in the prior art, and help ensure the efficient circulation and safe supply of medical supplies in all links of the supply chain.

[0039] Other beneficial effects of the present invention will be further described in the specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a flow chart of a medical digital supply chain risk analysis method in an embodiment;

[0041] Figure 2 This is a module block diagram of a medical digital supply chain risk analysis system in an embodiment;

[0042] Figure 3 It is a module block diagram of an electronic device in an embodiment. DETAILED DESCRIPTION

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0044] Embodiment 1:

[0045] This embodiment discloses a medical digital supply chain risk analysis method, which can be executed by, but is not limited to, a computer device or a virtual machine with certain computing resources, such as a personal computer, a smart phone, a personal digital assistant, or a wearable device, or by a virtual machine.

[0046] like Figure 1 As shown, a medical digital supply chain risk analysis method may include but is not limited to the following steps:

[0047] S1. Obtain the initial attribute information of each participant node in the specified supply chain, and construct a supply chain graph with each participant node in the specified supply chain as an entity and the connection relationship between each participant node in the specified supply chain as an edge. The initial attribute information of each participant node is H (0)’ = {I v ,D v ,T v}, including node inventory I v , Node demand D v and node transport status information T v , where the node inventory I v Used to indicate the current inventory level, node demand D v Used to indicate the current order demand, node transportation status information T v Including transportation delays and cargo damage rates, etc.; the initial attribute information between the nodes of each participant, such as the transportation time T i,j and transportation risk R i,j , where the transportation time T i,j Used to indicate that the goods are from the participant node v i Transmitted to the participant node v j Time required, transportation risk R i,j Used to indicate that the goods are from the participant node v i Transmitted to the participant node v j The risk level such as temperature deviation or humidity exceeding the standard can be determined.

[0048] In step S1, the supply chain graph is G = (V, E), where V = {v 1 ,v 2 ,...,v n},v 1 ,v 2 ,...,v n represents the nodes of each participant in the specified supply chain, including vaccine production plants, logistics storage centers, distribution stations, hospitals, etc., E = {e ij |v i ,v j ∈V}, indicating the connection relationships between the nodes of each participant, such as logistics paths and transaction relationships.

[0049] It should be noted that the supply chain diagram is an abstract modeling of the actual medical supply chain, providing an intuitive representation of supply chain nodes and their relationships, providing a structured basis for subsequent risk propagation modeling, and facilitating subsequent supply chain risk analysis. In addition, the supply chain diagram can support dynamic updates (such as adding new nodes and adjusting edge weights) to adapt to real-time changing supply chain scenarios.

[0050] S2. Collect multi-source heterogeneous data of the designated supply chain in real time, and perform data fusion processing on the multi-source heterogeneous data to obtain fused data.

[0051] In step S2, the multi-source heterogeneous data includes logistics data D IoT , purchasing data D ERP and external environment data D ext , the fused data is:

[0052] D fused =w 1 D IoT +w 2 D ERP +w 3 D ext ;

[0053] In the formula, w 1 is the weight of logistics data, w 2 is the weight of the purchase data, w 3 is the weight of external environment data, w 1 、w 2 and w 3 Can be adjusted dynamically based on historical data relevance.

[0054] As an example, D IoT ={x i,t}, represents the logistics data of the i-th medical supplies at time t, D ERP ={s i,t}, represents the purchase order data of the i-th medical supplies at time t, D ext ={e j,t}, represents the external environment data at time t.

[0055] It should be noted that, in this embodiment, based on the data fusion process, the system's ability to process multi-dimensional information can be improved, making the description of the supply chain status more comprehensive. By introducing weight coefficients, the influence of different data sources can be flexibly adjusted according to actual scenarios.

[0056] S3. Acquire multiple fused data of the specified supply chain at k historical time steps before the current moment to obtain a fused data set, and input the fused data set into a preset time series prediction model to obtain a time series prediction result.

[0057] In step S3, the fused data set is represented as: t ={D fused,t-k , D fused,t-k+1 , ..., D fused,t}; where k is the time window of the fused data set, D fused,t-k , D fused,t-k+1 , ..., D fused,t is the fused data at time tk, t-k+1, ... and t in the fused data set;

[0058] In step S3, the Transformer-based time series model is used, and the time series prediction results obtained are: In the formula, X t is the fused data set, and θ is a preset model parameter. In this embodiment, the time series prediction results include but are not limited to future demand, transportation bottleneck probability, inventory change information, etc., which are not limited here.

[0059] In this embodiment, the Transformer-based time series prediction model has a powerful ability to model long-term dependencies and can predict future risks more accurately.

[0060] S4. Based on the time series prediction results, the initial attribute information of each participant node in the specified supply chain is updated to obtain the updated attribute information of each participant node. It should be noted that the time series prediction results provide information such as the demand, inventory change information and transportation bottleneck probability of each node in the future period, which is used to dynamically update the attribute information of each participant node in the supply chain diagram so that the supply chain diagram reflects the future supply chain risk characteristics. In this embodiment, based on the time series prediction results, the initial attribute information of each participant node in the specified supply chain is updated, that is, the time series prediction results are integrated into the initial attribute information of each participant node. In this embodiment, the updated attribute information of each participant node is represented as

[0061] S5. Based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network (GCN) is used to obtain the risk score of each participant node in the supply chain graph. It should be noted that the risk score of each participant node is used to indicate the risk level of the corresponding participant. In this embodiment, after obtaining the risk score of each participant node, the global risk score of the specified supply chain can also be obtained by averaging to reflect the overall risk status of the specified supply chain.

[0062] In step S5, based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network is used to obtain a risk score of each participant node in the supply chain graph, including:

[0063] S501. Obtain an adjacency matrix matching the supply chain graph; specifically, in the implementation process, after the supply chain graph is constructed, the supply chain graph is represented by an adjacency matrix to quickly determine whether there is a connection relationship between any two participating nodes in the supply chain graph, and to facilitate subsequent operations; wherein, if a node v in the supply chain graph i and v j If there is a connection between them, then the element A in the adjacency matrix ij =1, otherwise A ij =0.

[0064] S502. Using the adjacency matrix and the updated attribute information of each participant node as input data of the graph convolutional network, so as to update the risk characteristics of the participant nodes in the supply chain graph through the graph convolutional network, and then obtain the node characteristics of each participant node in the supply chain graph;

[0065] In step S502, the formula for updating the risk characteristics of the participant nodes in the supply chain graph through the graph convolutional network is as follows:

[0066] H (l+1) =σ(AH (l) W (l) );

[0067] Where A is the adjacency matrix matching the supply chain graph, H (l) is the node feature output by the lth network layer in the graph convolutional network, l is a natural number greater than or equal to 0, H (l+1) is the node feature output by the l+1th network layer in the graph convolutional network, W (l) is the weight matrix of the lth network layer in the graph convolutional network, which is the parameter learned during the training of the graph convolutional network, and σ is a preset activation function, such as the ReLU (linear rectification function) activation function.

[0068] Specifically, in step S502 of this embodiment, the updated attribute information of each participant node is first input into the 0th network layer of the graph convolutional network; at the same time, the interaction between the participant nodes in the supply chain graph is calculated through the adjacency matrix to realize node feature propagation. For example, the participant node v 2 The shortage of inventory will be transmitted through the logistics path e 2,3 Propagate to the participating nodes v 3 , resulting in the participant node v 3 The risk of supply chain shortage increases; based on this, the graph convolutional network is updated layer by layer, and multi-layer feature extraction is performed to capture more complex upstream and downstream relationship influences, thereby obtaining the node features of the nodes of each participant in the supply chain graph, that is, the node features of the nodes of each participant output by the last network layer in the graph convolutional network, which include risk-related information such as inventory gaps and transportation bottleneck probabilities.

[0069] S503. According to the node characteristics of each participant node, the risk score of each participant node is obtained through a preset linear mapping function.

[0070] In this embodiment, after obtaining the risk score of each participant node, these node features are mapped into node risk scores through a preset linear mapping function f, which is used to characterize the risk level of each participant node in the specified supply chain. If the number of layers of the last network layer of the graph convolutional network is L, then any participant node v i The risk score of can be expressed as: S vi =f(H vi (L) ).

[0071] It should be noted that the graph convolutional network can capture the topological relationship in the supply chain graph and the risk propagation effect between nodes. This embodiment realizes the quantification of the risk scores of the nodes participating in the supply chain through the above scheme, which can provide a basis for subsequent high-risk node screening and optimization strategies.

[0072] In this embodiment, the graph convolutional network can enhance the ability to express complex network relationships and multi-dimensional features, and can more accurately identify the risk propagation pattern between nodes. For example, when logistics bottlenecks and insufficient inventory occur at the same time, multiple features can be combined to identify potential high-risk nodes.

[0073] S6. Filter out the participant nodes whose risk scores are greater than the preset risk threshold from the participant nodes, and construct a high-risk participant list based on the participant nodes whose risk scores are greater than the risk threshold. Specifically, in this embodiment, in the high-risk participant list, the high-risk participant nodes are sorted in descending order of risk scores, so that the administrator user can conduct risk screening on the participants in the specified supply chain, which helps decision makers quickly identify and prioritize key issues in the supply chain. The high-risk participant list is represented as: V risk = {v i |S vi >τ}, where τ is the preset risk threshold.

[0074] In step S5, after obtaining the list of high-risk participants, the method further includes:

[0075] S7. Based on the list of high-risk participants, a preset reinforcement learning model is used to generate a supply chain optimization strategy. The supply chain optimization strategy may include logistics route adjustment (replanning logistics routes to avoid transportation bottlenecks), inventory transfer (transferring inventory from low-risk participant nodes to high-risk participant nodes) and resource allocation (increasing transportation resources, accelerating production, etc.), and the supply chain optimization strategy is applied to the designated supply chain, such as replanning vehicle routes or reallocating inventory, etc., so as to conduct a feedback loop, form a state optimization closed loop, and ensure the overall efficiency of the supply chain and reduce risks.

[0076] This embodiment can realize risk monitoring and early warning of the supply chain with high real-time performance and accuracy. Specifically, during the implementation of this embodiment, the initial attribute information of each participant node in the specified supply chain is obtained in advance, and the node of each participant in the specified supply chain is taken as an entity, and the connection relationship between the nodes of each participant in the specified supply chain is taken as an edge to construct a supply chain graph; multi-source heterogeneous data of the specified supply chain is collected in real time, and the multi-source heterogeneous data is subjected to data fusion processing to obtain fused data, and then multiple fused data of the specified supply chain at k historical time steps before the current moment are obtained to obtain a fused data set, and the fused data set is input into a preset time series prediction model to obtain a time series prediction result; then, according to the time series prediction result, the initial attribute information of each participant node in the specified supply chain is updated to obtain the updated attribute information of each participant node; then, based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network is used to obtain the risk score of each participant node in the supply chain graph; finally, from each participant node, a participant node with a risk score greater than a preset risk threshold is screened, and a high-risk participant list is constructed based on the participant node with a risk score greater than the risk threshold. In this embodiment, the supply chain graph and time series prediction model are used to dynamically update node attribute information and perform subsequent calculations on the risk scores of participating nodes, which can help to comprehensively capture the dynamic changes in the specified supply chain structure, enhance the sensitivity of risk warning, and meet the needs of real-time monitoring and rapid response to the entire supply chain process; in addition, the introduction of the graph convolutional network setting in risk propagation modeling can effectively identify potential risk paths, solve the problem of difficulty in risk identification in complex scenarios in existing technologies, and help ensure the efficient circulation and safe supply of medical supplies in all links of the supply chain.

[0077] Embodiment 2:

[0078] This embodiment discloses a medical digital supply chain risk analysis system for implementing the medical digital supply chain risk analysis method in Embodiment 1; Figure 2 As shown, the medical digital supply chain risk analysis system includes:

[0079] A graph construction module is used to obtain initial attribute information of each participant node in a specified supply chain, and construct a supply chain graph by taking each participant node in the specified supply chain as an entity and the connection relationship between each participant node in the specified supply chain as an edge;

[0080] An attribute updating module is communicatively connected with the graph building module, and is used to collect multi-source heterogeneous data of the specified supply chain in real time, and perform data fusion processing on the multi-source heterogeneous data to obtain fused data; is used to obtain multiple fused data of the specified supply chain at k historical time steps before the current moment to obtain a fused data set, and input the fused data set into a preset time series prediction model to obtain a time series prediction result; and is also used to update the initial attribute information of each participant node in the specified supply chain according to the time series prediction result to obtain the updated attribute information of each participant node;

[0081] A risk score acquisition module, which is in communication with the attribute update module and is used to obtain the risk score of each participant node in the supply chain graph using a graph convolutional network based on the supply chain graph and the updated attribute information of each participant node;

[0082] The node screening module is communicatively connected with the risk score acquisition module, and includes screening out the participant nodes whose risk scores are greater than a preset risk threshold from the participant nodes, and constructing a high-risk participant list based on the participant nodes whose risk scores are greater than the risk threshold.

[0083] It should be noted that the working process, working details and technical effects of the medical digital supply chain risk analysis system provided in this Example 2 can be found in Example 1 and will not be repeated here.

[0084] Embodiment 3:

[0085] Based on Embodiment 1 or 2, this embodiment discloses an electronic device, which may be a smart phone, a tablet computer, a laptop computer, or a desktop computer. The electronic device may be referred to as a user terminal, a portable terminal, a desktop terminal, etc. Figure 3 As shown, the electronic equipment includes:

[0086] a memory for storing computer program instructions; and,

[0087] A processor is used to execute the computer program instructions to complete the operation of a medical digital supply chain risk analysis method as described in any one of Example 1.

[0088] Specifically, the processor 301 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 301 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 301 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 301 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen.

[0089] The memory 302 may include one or more computer-readable storage media, which may be non-transitory. The memory 302 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 302 is used to store at least one instruction, which is used to be executed by the processor 301 to implement the medical digital supply chain risk analysis method provided in Example 1 of the present application.

[0090] In some embodiments, the terminal may further optionally include: a communication interface 303 and at least one peripheral device. The processor 301, the memory 302 and the communication interface 303 may be connected via a bus or a signal line. Each peripheral device may be connected to the communication interface 303 via a bus, a signal line or a circuit board. Specifically, the peripheral device includes: at least one of a radio frequency circuit 304, a display screen 305 and a power supply 306.

[0091] The communication interface 303 may be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 301 and the memory 302. In some embodiments, the processor 301, the memory 302, and the communication interface 303 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 301, the memory 302, and the communication interface 303 may be implemented on a separate chip or circuit board, which is not limited in this embodiment.

[0092] The radio frequency circuit 304 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 304 communicates with a communication network and other communication devices through electromagnetic signals.

[0093] The display screen 305 is used to display a UI (User Interface). The UI may include any combination of graphics, text, icons, and videos.

[0094] The power supply 306 is used to supply power to various components in the electronic device.

[0095] Embodiment 4:

[0096] Based on any one of Embodiments 1 to 3, this embodiment discloses a computer program product, including a computer program or instructions, which, when executed by a computer, implements a medical digital supply chain risk analysis method as described in any one of Embodiments 1. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0097] Embodiment 5:

[0098] Based on any one of Embodiments 1 to 3, this embodiment discloses a computer-readable storage product, on which instructions are stored, and when the instructions are run on a computer, a medical digital supply chain risk analysis method as described in any one of Embodiments 1 is executed. The computer-readable storage product refers to a carrier for storing data, which may include but is not limited to computer-readable storage media such as floppy disks, optical disks, hard disks, flash memories, USB flash drives, and / or memory sticks, and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0099] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the above embodiments, a person skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A medical digital supply chain risk analysis method, characterized in that: include: Acquire initial attribute information of each participant node in the specified supply chain, and construct a supply chain graph with each participant node in the specified supply chain as an entity and the connection relationship between each participant node in the specified supply chain as an edge; Collect multi-source heterogeneous data of the specified supply chain in real time, and perform data fusion processing on the multi-source heterogeneous data to obtain fused data; Acquire multiple fused data of the specified supply chain at k historical time steps before the current moment to obtain a fused data set, and input the fused data set into a preset time series prediction model to obtain a time series prediction result; According to the time series prediction result, the initial attribute information of each participant node in the specified supply chain is updated to obtain updated attribute information of each participant node; Based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network is used to obtain a risk score of each participant node in the supply chain graph; From each participant node, the participant nodes whose risk scores are greater than a preset risk threshold are screened out, and based on the participant nodes whose risk scores are greater than the risk threshold, a high-risk participant list is constructed.

2. A medical digital supply chain risk analysis method according to claim 1, characterized in that: The supply chain graph is G = (V, E), where V = {v1, v2, ..., v n },v1,v2,...,v n represents the nodes of each participant in the specified supply chain, E = {e ij |v i ,v j ∈V}, indicating the connection relationship between the nodes of each participant.

3. A medical digital supply chain risk analysis method according to claim 1, characterized in that: The multi-source heterogeneous data includes logistics data D IoT , purchasing data D ERP and external environment data D ext , the fused data is: D fused =w1D IoT +w2D ERP +w3D ext ; In the formula, w1 is the weight of logistics data, w2 is the weight of procurement data, and w3 is the weight of external environment data.

4. A medical digital supply chain risk analysis method according to claim 1, characterized in that: The fused data set is represented as: t ={D fused,t-k , D fused,t-k+1 , ..., D fused,t }; where k is the time window of the fused data set, D fused,t-k , D fused,t-k+1 , ..., D fused,t is the fused data at time tk, t-k+1, ... and t in the fused data set; Using the Transformer-based time series model, the time series prediction results are: Where, X t is the fused data set, and θ is the preset model parameter.

5. A medical digital supply chain risk analysis method according to claim 1, characterized in that: Based on the supply chain graph and the updated attribute information of each participant node, a graph convolutional network is used to obtain the risk score of each participant node in the supply chain graph, including: Obtaining an adjacency matrix matching the supply chain graph; The adjacency matrix and the updated attribute information of each participant node are used as input data of a graph convolutional network, so as to update the risk characteristics of the participant nodes in the supply chain graph through the graph convolutional network, thereby obtaining the node characteristics of each participant node in the supply chain graph; According to the node characteristics of each participant's node, the risk score of each participant's node is obtained through a preset linear mapping function.

6. A medical digital supply chain risk analysis method according to claim 5, characterized in that: The formula for updating the risk characteristics of the participant nodes in the supply chain graph through the graph convolutional network is as follows: H (l+1) =σ(AH (l) W (l) ); Where A is the adjacency matrix matching the supply chain graph, H (l) is the node feature output by the lth network layer in the graph convolutional network, l is a natural number greater than or equal to 0, H (l+1) is the node feature output by the l+1th network layer in the graph convolutional network, W (l) is the weight matrix of the lth network layer in the graph convolutional network, and σ is the preset activation function.

7. A medical digital supply chain risk analysis system, characterized in that: include: A graph construction module is used to obtain initial attribute information of each participant node in a specified supply chain, and construct a supply chain graph by taking each participant node in the specified supply chain as an entity and the connection relationship between each participant node in the specified supply chain as an edge; An attribute updating module is connected to the graph building module for real-time collection of multi-source heterogeneous data of the specified supply chain, and performs data fusion processing on the multi-source heterogeneous data to obtain fused data; it is used to obtain multiple fused data of the specified supply chain at k historical time steps before the current moment to obtain a fused data set, and input the fused data set into a preset time series prediction model to obtain a time series prediction result; It is also used to update the initial attribute information of each participant node in the specified supply chain according to the time series prediction result to obtain the updated attribute information of each participant node; A risk score acquisition module, which is in communication with the attribute update module and is used to obtain the risk score of each participant node in the supply chain graph using a graph convolutional network based on the supply chain graph and the updated attribute information of each participant node; The node screening module is communicatively connected with the risk score acquisition module, and includes screening out the participant nodes whose risk scores are greater than a preset risk threshold from the participant nodes, and constructing a high-risk participant list based on the participant nodes whose risk scores are greater than the risk threshold.

8. An electronic device, characterized in that: include: a memory for storing computer program instructions; as well as, A processor, used to execute the computer program instructions to complete the operation of a medical digital supply chain risk analysis method as described in any one of claims 1 to 6.

9. A computer program product comprising a computer program or instructions, characterized in that When executed by a computer, the computer program or the instruction implements a medical digital supply chain risk analysis method as described in any one of claims 1 to 6.

10. A computer readable storage product, characterized in that: The computer-readable storage product stores instructions, and when the instructions are executed on a computer, a medical digital supply chain risk analysis method as described in any one of claims 1 to 6 is executed.

Images