A method, apparatus, equipment, and medium for optimizing a supply chain network.

Abstract

This invention belongs to the technical field of supply chain network optimization, and specifically relates to a method, apparatus, equipment, and medium for optimizing supply chain networks. The invention provides a supply chain network optimization method comprising: establishing an initial supply chain network with defined rapid response capabilities for each node; calculating the comprehensive centrality index of each node in the initial supply chain network; performing simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results; determining the optimized values ​​for each parameter in the initial supply chain network based on the simulation analysis results; and optimizing the initial supply chain network based on the optimized values ​​to obtain an optimized supply chain network. This effectively solves the problems of insufficient characterization of dynamic rapid response capabilities by traditional models in existing research and the limited application of deep learning in the field of emergency supply chains, promoting the application and development of complex network theory and deep learning in the field of emergency supply chains.

Description

Technical Field

[0001] This invention belongs to the technical field of supply chain network optimization, and specifically relates to a method, apparatus, equipment and medium for optimizing a supply chain network. Background Technology

[0002] Cascading failure refers to a phenomenon in a network system where a failure in one of two adjacent nodes can trigger a failure in the next adjacent node, and this propagation continues until the entire network collapses. Adilson E. Motter and Ying-Cheng Lai (2002) first systematically proposed the propagation mechanism of cascading failure in complex networks and introduced the concepts of node load and capacity. They simulated a scenario where load redistribution after a node failure leads to overload of other nodes, thus triggering cascading failure. This not only emphasized the crucial role of high-load nodes in network stability but also found that load-based deliberate attacks are more destructive than random attacks. Since then, cascading failure has rapidly become an important area of ​​research in complex networks.

[0003] Supply chain networks, as typical complex networks, are generally studied starting with building cascading failure models that conform to the specific characteristics of a particular industry. The failure process is then simulated, and the simulation results are discussed. While deep learning has made some progress in addressing cascading failures in emergency supply chains, many challenges and unresolved issues remain. For example, data acquisition and quality are critical issues, as data in emergency supply chains often exhibits sparsity, incompleteness, and noise interference. Furthermore, deep learning models suffer from poor interpretability, making it difficult to directly provide supply chain managers with intuitive risk analysis and decision-making support. Therefore, improving the robustness and interpretability of models, and better integrating deep learning techniques with other methods (such as physical models and optimization algorithms), remain important directions for future research. Summary of the Invention

[0004] This invention provides a supply chain network optimization method, comprising: establishing an initial supply chain network with defined rapid response capabilities for each node; calculating the comprehensive centrality index of each node in the initial supply chain network; performing simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results; determining the optimized values ​​for each parameter in the initial supply chain network based on the simulation analysis results; and optimizing the initial supply chain network based on the optimized values ​​to obtain an optimized supply chain network. This effectively solves the problems of insufficient characterization of dynamic rapid response capabilities by traditional models in existing research and the limited application of deep learning in the field of emergency supply chains. It provides a scientific basis and practical guidance for the optimization and resilience enhancement of emergency supply chains, and promotes the application and development of complex network theory and deep learning in the field of emergency supply chains.

[0005] To address the aforementioned technical problems, this application proposes four aspects.

[0006] In a first aspect, this application provides a method for optimizing a supply chain network, comprising: establishing an initial supply chain network for which rapid response capabilities are defined for each node; calculating the comprehensive centrality index of each node in the initial supply chain network; performing simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results; determining the optimization values ​​for each parameter in the initial supply chain network based on the simulation analysis results; and optimizing the initial supply chain network based on the optimization values ​​to obtain an optimized supply chain network.

[0007] In some embodiments, the step of simulating and analyzing the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results includes: identifying nodes whose comprehensive centrality index meets preset conditions as target nodes; performing a simulated attack on the target nodes; setting different values ​​for each parameter of the initial supply chain network during the simulation attack; calculating the cascading failure index of each parameter during the simulation attack; and generating multiple cascading failure index-parameter data tables based on the cascading failure index of each parameter.

[0008] In some embodiments, determining the optimized values ​​of each parameter in the initial supply chain network based on the simulation analysis results includes: using the parameter value corresponding to the maximum cascading failure index in each of the cascading failure index-parameter data tables as the optimized value.

[0009] In some embodiments, establishing an initial supply chain network with defined rapid response capabilities for each node includes: defining the rapid response capability of each node; determining the initial load of each node based on the rapid response capability and a rapid response capability adjustment coefficient; determining the node capacity of each node based on the initial load and the rapid response capability of the node; and determining the load redistribution rules of the initial supply chain network based on the rapid response capability of each node.

[0010] In some embodiments, establishing an initial supply chain network with defined rapid response capabilities for each node further includes: determining the initial load of each node using the following formula: 1+ )in, Let be the initial load of node i. λ is the degree value of node i; x represents the load index; λ is the adjustment coefficient of the node's fast response capability, used to adjust the impact of fast response capability on the initial load; This represents the fast response capability of node i; the node capacity includes an upper capacity limit and a lower capacity limit, wherein the upper capacity limit is determined by the following formula: Let β represent the upper limit of capacity, β represent the capacity growth factor, and μ represent the adjustment coefficient of the rapid response capability on the upper limit of capacity; the lower limit of capacity is determined by the following formula: in, Here, y represents the lower limit of capacity, and ν represents the adjustment coefficient of the fast response capability to the lower limit of capacity; the load redistribution rules are as follows: in, This represents the load that was distributed from the failed node to its neighboring nodes; , where is the remaining capacity of neighbor node j. Mark the capacity limit of neighbor node j. This indicates the current load of neighbor node j; η Γi represents the adjustment coefficient of rapid response capability to load distribution, and Γi represents the set of neighboring nodes of the failed node i. This indicates the fast response capability of neighbor node j. Li This represents the load of the failed node i. To represent a constant and prevent multiplication by zero; for the neighbor node j after receiving the load redistributed by the failed node i, the load of neighbor node j is: In the formula, The current load of neighbor node J after it has accepted the redistributed load is given. , , This represents the load received by node j. This represents the load fluctuation caused by node j after it accepts the load. It is a factor that controls the amplitude of fluctuations.

[0011] In some embodiments, the method further includes: combining a DNN model with a LightGBM model to form an initial ensemble model; optimizing the parameters of the initial ensemble model using a Nadam optimizer to form an optimized ensemble model; using the optimized ensemble model to predict the nodes of the optimized supply chain network during its operation; and optimizing the load of nodes at risk based on the prediction results.

[0012] In some embodiments, the parameters that need to be optimized for the initial supply chain network include: the node's rapid response capability, the adjustment coefficient of the node's influence capability, the capacity growth factor, the adjustment coefficient of the capacity ceiling, the capacity floor parameter, the adjustment coefficient of the capacity floor, the adjustment coefficient of the rapid response capability on load distribution, and the factor for controlling the fluctuation amplitude.

[0013] Secondly, this application proposes a supply chain network optimization device, comprising: a first execution module for establishing an initial supply chain network with defined rapid response capabilities for each node; a first calculation module for calculating the comprehensive centrality index of each node in the initial supply chain network; a second execution module for performing simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results; a first determination module for determining the optimization values ​​of each parameter in the initial supply chain network based on the simulation analysis results; and a first optimization module for optimizing the initial supply chain network based on the optimization values ​​to obtain an optimized supply chain network.

[0014] Thirdly, this application proposes a computer electronic production apparatus, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any of the methods described in the first aspect.

[0015] Fourthly, this application proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any one of the first aspects of the claims.

[0016] This invention provides a supply chain network optimization method, comprising: establishing an initial supply chain network with defined rapid response capabilities for each node; calculating the comprehensive centrality index of each node in the initial supply chain network; performing simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results; determining the optimized values ​​for each parameter in the initial supply chain network based on the simulation analysis results; and optimizing the initial supply chain network based on the optimized values ​​to obtain an optimized supply chain network. This effectively solves the problems of insufficient characterization of dynamic rapid response capabilities by traditional models in existing research and the limited application of deep learning in the field of emergency supply chains. It provides a scientific basis and practical guidance for the optimization and resilience enhancement of emergency supply chains, and promotes the application and development of complex network theory and deep learning in the field of emergency supply chains. Attached Figure Description

[0017] One or more embodiments are illustrated by way of example with reference to the accompanying drawings, and these illustrative descriptions do not constitute a limitation on the embodiments.

[0018] Figure 1 The main flowchart of a supply chain network optimization method provided in this application embodiment;

[0019] Figure 2 A diagram showing the influence of capacity coefficient β and response coefficient μ on the cascade failure index provided in the embodiments of this application;

[0020] Figure 3 A diagram showing the relationship between the load index x and the response weight λ on the cascade failure index, provided in an embodiment of this application.

[0021] Figure 4 A main structural block diagram of a supply chain network optimization device provided in an embodiment of this application;

[0022] Figure 5 This is a structural block diagram of a computer electronic production equipment provided in an embodiment of this application. Detailed Implementation

[0023] Cascading failure refers to a phenomenon in a network system where a failure in one of two adjacent nodes can trigger a failure in the next adjacent node, and this propagation continues until the entire network collapses. Adilson E. Motter and Ying-Cheng Lai (2002) first systematically proposed the propagation mechanism of cascading failure in complex networks and introduced the concepts of node load and capacity. They simulated a scenario where load redistribution after a node failure leads to overload of other nodes, thus triggering cascading failure. This not only emphasized the crucial role of high-load nodes in network stability but also found that load-based deliberate attacks are more destructive than random attacks. Since then, cascading failure has rapidly become an important area of ​​research in complex networks.

[0024] Supply chain networks, as typical complex networks, are generally studied starting with building cascading failure models that conform to the specific characteristics of a particular industry. The failure process is then simulated, and the simulation results are discussed. While deep learning has made some progress in addressing cascading failures in emergency supply chains, many challenges and unresolved issues remain. For example, data acquisition and quality are critical issues, as data in emergency supply chains often exhibits sparsity, incompleteness, and noise interference. Furthermore, deep learning models suffer from poor interpretability, making it difficult to directly provide supply chain managers with intuitive risk analysis and decision-making support. Therefore, improving the robustness and interpretability of models, and better integrating deep learning techniques with other methods (such as physical models and optimization algorithms), remain important directions for future research.

[0025] To address the aforementioned technical problems, this invention proposes a supply chain network optimization method. The implementation details of the bandwidth determination method in this embodiment are described below. The following content is only for ease of understanding and is not necessary for implementing this solution.

[0026] Example 1:

[0027] like Figure 1 As shown, this application provides a method for optimizing a supply chain network. This method is applicable to electronic production equipment, which can be a server, mobile terminal, computer, cloud platform, etc. The functions implemented by the data processing of the production equipment provided in this application embodiment can be achieved by the processor of the electronic production equipment calling program code, wherein the program code can be stored in a computer storage medium. The method for optimizing the supply chain network includes:

[0028] Step S1: Establish an initial supply chain network with defined rapid response capabilities for each node.

[0029] By defining the rapid response capability, initial load, capacity upper and lower limits, and state recognition mechanism of nodes, this application constructs an initial supply chain network that considers rapid response capability. In this initial supply chain network, this application sets the relationship between various attributes and rapid response capability, and introduces several new parameters. The subsequent optimization of the initial supply chain network in this application is to optimize these parameters in order to improve the robustness of the supply chain network, so that the optimized supply chain network can have the best robustness.

[0030] Therefore, in some embodiments, step S1, "establishing an initial supply chain network with defined rapid response capabilities for each node," includes:

[0031] Step S11: Define the fast response capability of each node.

[0032] Rapid response capability indicates the degree of load change that a node can handle per unit time, simulating the node's ability to respond quickly. High-response nodes are top 10% degree central nodes, with an R (rapid response capability) of 0.7-1; low-response nodes have an R of 0.01-0.3.

[0033] Step S12: Determine the initial load of each node based on the node's fast response capability and fast response capability adjustment coefficient.

[0034] In emergency supply chains, the initial load of a node depends not only on its degree value but also on its rapid response capability. Nodes with strong rapid response capabilities can process tasks more efficiently, therefore their initial load can be appropriately increased, as follows:

[0035] 1+ )

[0036] in, Let be the initial load of node i. Let be the degree value of node i. Let x represent the load index. λ is the adjustment coefficient for the node's fast response capability, used to adjust the impact of fast response capability on the initial load. This represents the fast response capability of node i (e.g., the ability to process tasks per unit of time).

[0037] Step S13: Determine the node capacity of each node based on the initial load and the node's fast response capability.

[0038] Node capacity includes an upper and a lower limit, and the impact of rapid response capability should be taken into account. Nodes with strong rapid response capability can utilize their capacity more effectively in emergency situations, so the upper limit can be appropriately increased, while the lower limit can be appropriately decreased to reflect their flexibility.

[0039] The upper limit of capacity is determined by the following formula:

[0040]

[0041] β represents the capacity limit, μ represents the capacity growth factor, and μ represents the adjustment coefficient of the rapid response capability on the capacity limit.

[0042] The lower limit of capacity is determined by the following formula:

[0043]

[0044] in, Here, y represents the lower limit of capacity, and ν represents the adjustment coefficient of the rapid response capability on the lower limit of capacity.

[0045] If the load exceeds the capacity limit, the node enters an abnormal state, and the duration of the abnormality is accumulated: if the node remains abnormal for a longer period of time than [a certain duration], the node will enter an abnormal state. If the duration is less than 100, then it will fail. So, reducing response capability, i.e. A grace period is set for nodes to allow for recovery after brief fluctuations; the node failure threshold is defined as follows. The node failure time is: .

[0046] Step S14: Determine the load redistribution rules of the initial supply chain network based on the rapid response capabilities of each node.

[0047] After a node fails, it is necessary to quickly redistribute the load of the failed node to other nodes, while ensuring that other nodes do not crash after receiving the redistributed load. Therefore, it is necessary to define load redistribution rules, which are described in this application as follows:

[0048]

[0049] in, This represents the load that was distributed from the failed node to the neighboring nodes. , where is the remaining capacity of neighbor node j. Mark the capacity limit of neighbor node j. This indicates the current load of neighbor node j. η Γi represents the adjustment coefficient of rapid response capability to load distribution, and Γi represents the set of neighboring nodes of the failed node i. This indicates the fast response capability of neighbor node j. Li This represents the load of the failed node i. This represents a constant to prevent errors such as multiplying by 0.

[0050] After receiving the load redistributed by the failed node i, neighbor node j needs to update its load. After the update, the load of neighbor node j is:

[0051]

[0052] In the formula, The current load of neighbor node J after it has accepted the redistributed load is given. , , This represents the load received by node j. This represents the load fluctuation caused by node j after it accepts the load. It is a factor that controls the amplitude of fluctuations.

[0053] From the above-described initial supply chain network settings, it can be seen that this application has many more parameters that need to be adjusted. These parameters include: node fast response capability (R), fast response capability adjustment coefficient (λ), capacity growth factor (β), capacity ceiling adjustment coefficient (μ), capacity floor parameter (y), capacity floor adjustment coefficient (v), and fast response capability adjustment coefficient for load allocation. η ) and factors that control the amplitude of fluctuations ( These eight parameters are also the parameters that need to be numerically optimized later. By optimizing the values ​​of these parameters, the optimized supply chain network will have better robustness.

[0054] Step S2: Calculate the comprehensive centrality index of each node in the initial supply chain network.

[0055] To optimize the initial supply chain network, it is necessary to first conduct simulation analysis of the initial supply chain network. This simulation analysis needs to be based on a simulated attack on the supply chain network. However, a simulated attack on the supply chain network does not attack all network nodes, but generally attacks the central nodes of the network. In general, a comprehensive centrality index can determine which nodes are central nodes. Therefore, this application also needs to calculate the comprehensive centrality index of each node.

[0056] The overall centrality index of each node is calculated using the following formula:

[0057]

[0058] Degree Centrality calculates the number of other nodes directly connected to a given node; Betweenness refers to the proportion of shortest paths in the network that pass through that node; Coreness refers to the degree of centrality of a node in the network.

[0059] Step S3: Perform simulation analysis on the initial supply chain network based on the comprehensive centrality index, and obtain the simulation analysis results.

[0060] In some embodiments, step S3, "simulation analysis of the initial supply chain network based on the comprehensive centrality index, and obtaining simulation analysis results," includes:

[0061] Step S31: Nodes whose comprehensive centrality index meets the preset conditions are identified as target nodes.

[0062] Since the simulation analysis process requires simulating attacks on centrality nodes, this application identifies the nodes with a comprehensive centrality index in the top 10% as the target nodes for simulation attacks.

[0063] Step S32: Perform a simulated attack on the target node.

[0064] Step S33: During the simulation attack, set different values ​​for each parameter of the initial supply chain network.

[0065] Step S34: Calculate the cascading failure index of each parameter during the simulated attack process.

[0066] Step S35: Generate multiple cascade failure index-parameter data tables based on the cascade failure indices of each parameter.

[0067] Before conducting the simulated attack, multiple initial supply chain network replicas are created. These initial supply chain network replicas are assigned different values ​​to the eight parameters mentioned above. The purpose of the simulated attack is to determine the relationship between the parameters and the network robustness. The cascading failure index is an indicator for evaluating the network robustness. Therefore, during the simulated attack, it is necessary to calculate the cascading failure index of each initial supply chain network replica under each parameter, thereby obtaining the cascading failure index of the same parameter with different values ​​in the simulated attack. This allows the formation of a cascading failure index-parameter data table, which stores the cascading failure index corresponding to different values ​​of the same parameter.

[0068] For example, in this application, the formula for calculating the cascade failure index for β and μ is:

[0069] , where mean(d) is the degree average, and 0.1 is used to prevent the denominator from being 0.

[0070] For x and λ, the formula for calculating the cascade failure index is:

[0071] ,in The standard deviation is denoted as .

[0072] Of course, the above only provides the calculation formula for the cascading failure index with two sets of parameters. In fact, when calculating the cascading failure index, this application will select two parameters as a set to calculate the cascading failure index. The calculation formula for the cascading failure index will be changed according to the relationship between the two parameters in the same set. The specific change method is to substitute the relationship between the two parameters in the same set into the initial calculation formula for the cascading failure index, and then obtain the calculation formula for the cascading failure index that meets the parameters to be calculated.

[0073] Figure 2 This indicates the influence of the capacity coefficient β and the response coefficient μ on the cascade failure exponent, derived from... Figure 2 It is evident that the risk is highest when μ increases and β decreases. This indicates that the system is highly vulnerable when nodes are overly reliant on response values ​​and have insufficient remaining capacity. Figure 3 The graph illustrates the impact of the load index x and the response weight λ on the cascading failure index. As can be seen from the graph, when both λ and x increase simultaneously, the IgCI exhibits a sharp upward trend. This indicates that uneven load distribution and high response weights significantly increase the risk of network collapse. In emergency supply chain risk management, when the load on core nodes is excessively concentrated and their sensitivity to external interference is too high, the system has a very high probability of experiencing cascading failures.

[0074] Step S4: Determine the optimized values ​​of each parameter in the initial supply chain network based on the simulation analysis results.

[0075] In some embodiments, step S4, "determining the optimized values ​​for each parameter in the initial supply chain network based on the simulation analysis results," includes:

[0076] Step S41: The optimized value is the parameter value corresponding to the maximum cascading failure index in each of the cascading failure index-parameter data tables.

[0077] Therefore, this application can generate multiple cascading failure index-parameter data tables in the simulation attack. If this application wants to obtain the supply chain network with optimal robustness, it needs to determine the corresponding parameter values ​​in multiple cascading failure index-parameter data tables. Generally, the value corresponding to the largest cascading failure index in each cascading failure index-parameter data table is selected as the optimization value.

[0078] Step S5: Optimize the initial supply chain network based on the optimization values ​​to obtain an optimized supply chain network.

[0079] Therefore, this application introduces "rapid response capability" as a core indicator for nodes, and classifies nodes into high-response nodes and low-response nodes based on degree centrality. Subsequent node initial load, capacity upper and lower limits, and load redistribution processes are dynamically adjusted based on their response capability. Simulation results demonstrate the impact of node response capability on its failure probability; that is, the higher the node response capability, the lower the failure probability. This mechanism innovatively simulates the significant influence of critical node response capability on the entire network under real-world emergency scenarios, thereby enhancing the practical explanatory power of network resilience analysis.

[0080] This application incorporates the rapid processing capabilities of nodes by defining their responsiveness, enabling the network to more accurately simulate node behavior in the face of sudden load changes. The network details how the load of a failed node is redistributed to its neighboring nodes, taking into account load fluctuations and dynamic changes in node capacity. By defining metrics such as the Inverse Failure Index (IgCI), the network can quantify its robustness, providing an effective tool for assessing the resilience of supply chain networks.

[0081] In addition to optimizing internal network parameters to improve network performance, optimizing each node during network operation can further enhance network stability.

[0082] Therefore, in some embodiments, the method further includes:

[0083] Step S61: Combine the DNN model with the LightGBM model to form an initial ensemble model.

[0084] Step S62: Optimize the parameters of the initial ensemble model using the Nadam optimizer to form an optimized ensemble model.

[0085] Simulation analysis reveals the mechanism by which node response capability affects network robustness, while model ensemble aims to further enhance the model's predictive and generalization abilities. By combining DNN and LightGBM, and utilizing the leaf node paths of LightGBM as additional features input to the neural network, a more efficient and accurate prediction model is constructed. The LightGBM algorithm in machine learning can efficiently handle large-scale datasets, providing fast training and high accuracy, while supporting the handling of categorical features and preventing overfitting. Therefore, this paper integrates the DNN model with the LightGBM model using an optimizer, extracting the leaf node paths of the LightGBM model as features. The concatenated original features and tree node path features are input into the DNN model, and grid search is used to fine-tune the neural network model. The DNN has already been optimized using Nadam. The leaf node paths of LightGBM provide new perspectives and information, which can help the DNN better understand the data. The Nadam optimizer can more efficiently adjust model parameters, while grid search finds the most suitable configuration by trying different parameter combinations, thereby improving the model's performance and generalization ability, ultimately resulting in an optimized ensemble model.

[0086] Step S63: During the operation of the optimized supply chain network, the optimized integrated model is used to perform node prediction for each node of the optimized supply chain network.

[0087] Step S64: Optimize the load of nodes at risk based on the prediction results of the nodes.

[0088] By optimizing the ensemble model to acquire features of each node during the operation of the supply chain network, calculating and ranking the SHAP values ​​of these features based on their importance, and generating a SHAP-based beehive graph, the beehive graph is used to determine the prediction results for each node. Some predictions may indicate that the corresponding node is at risk of failure, thus identifying nodes at risk of failure as nodes to be optimized. Then, based on the current load of each node to be optimized, a load optimization strategy is determined for each node, thereby completing the load optimization of the operating supply chain network and improving its stability. This scheme effectively integrates the strong structural expressive power of tree models with the superior generalization ability of deep learning, providing an application tool for accurately identifying critical failure nodes and improving risk warning and intervention decisions in emergency supply chain networks.

[0089] It is worth noting that the combination of complex networks and deep learning is not a simple "patchwork," but rather a complementary and reinforcing relationship. Complex networks provide deep learning with structured data and prior knowledge, while deep learning provides complex networks with powerful data processing and predictive capabilities. In research on cascading failures in emergency supply chains, this combination can fully leverage the advantages of both, improving the predictive power and robustness of the model, and providing strong support for the optimization and resilience enhancement of emergency supply chains. Introducing deep learning for prediction is not intended to replace physical models, but rather to compensate for their shortcomings and improve their predictive power and adaptability. By combining physical models with deep learning, the advantages of both can be fully utilized to achieve more accurate, efficient, and reliable prediction and risk management of cascading failures in emergency supply chain networks. The approach of generating data from physical models and using it as input for feature selection and prediction in deep learning is an efficient and complementary research method that fully leverages the prior knowledge of physical models and the data-driven advantages of deep learning. Physical models, based on the physical laws and mathematical descriptions of the system, can generate high-quality simulation data with a clear physical background, covering various network states and failure modes. This data is not only of high quality but also rich in diversity, providing abundant learning material for deep learning models.

[0090] Therefore, this application incorporates the rapid processing capabilities of nodes into the definition of node responsiveness, enabling the network to more accurately simulate node behavior in the face of sudden load changes. The network details how the load of a failed node is redistributed to its neighboring nodes, considering load fluctuations and dynamic changes in node capacity. By defining metrics such as the Cascading Failure Index (IgCI), the network's robustness can be quantified, providing an effective tool for assessing the resilience of supply chain networks. By constructing a cascading failure network that considers node responsiveness, the dynamic evolution mechanism of the network is clarified. Furthermore, an ensemble model generated using DNN and LightGBM models is used to further explore the complex relationships between features, significantly improving the accuracy of cascading failure prediction and the generalization ability of the ensemble model. In addition, this study optimizes key parameters through simulation analysis, providing a scientific basis and practical guidance for the optimization and resilience enhancement of emergency supply chains, and promoting the application and development of complex network theory and deep learning in the field of emergency supply chains.

[0091] Example 2:

[0092] Based on the foregoing embodiments, this application provides a supply chain network optimization device. The various modules and units included in the device can be implemented by a processor in a computer device; of course, they can also be implemented by specific logic circuits. In the implementation process, the processor can be a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP), or a field programmable gate array (FPGA), etc.

[0093] like Figure 4 As shown, a supply chain network optimization device includes: a first execution module 1, a first calculation module 2, a second execution module 3, a first determination module 4, and a first optimization module 5.

[0094] The first execution module 1 is used to establish an initial supply chain network with defined rapid response capabilities for each node. The first calculation module 2 is used to calculate the comprehensive centrality index of each node in the initial supply chain network. The second execution module 3 is used to perform simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtain the simulation analysis results. The first determination module 4 is used to determine the optimized values ​​for each parameter in the initial supply chain network based on the simulation analysis results. The first optimization module 5 is used to optimize the initial supply chain network based on the optimized values ​​to obtain an optimized supply chain network.

[0095] The modules in the aforementioned supply chain network optimization device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor within the device in hardware form, or stored in the memory of the processing device in software form, so that the processor can call and execute the operations corresponding to each module. It should be noted that the module division in this embodiment is illustrative and represents only a logical functional division; in actual implementation, other division methods may be used.

[0096] Example 3:

[0097] Thirdly, this application provides a computer electronic production device, such as... Figure 5 As shown, it includes: at least one processor 901; and a memory 902 communicatively connected to the at least one processor 901; wherein the memory 902 stores instructions executable by the at least one processor 901, the instructions being executed by the at least one processor 901 to enable the at least one processor 901 to execute a drilling early warning method in the above embodiments.

[0098] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0099] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0100] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0101] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0102] In addition, the computer device may include (but is not limited to) a data bus, an input / output (I / O) bus, a display, and input / output devices (e.g., keyboard, mouse, speakers, etc.).

[0103] The processor can communicate with external devices via the I / O bus through wired or wireless networks.

[0104] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product / computer program product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.

[0105] Example 4:

[0106] Fourthly, this application proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the first aspects.

[0107] Computer-readable storage media can be implemented by any type of volatile or non-volatile storage device or a combination thereof. Computer-readable storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, and computer storage media (e.g., hard disks, floppy disks, solid-state drives, removable disks, CD-ROMs, DVD-ROMs, Blu-ray discs, etc.).

[0108] Computer-readable storage media may also store at least one computer-executable program / instruction, such as computer-readable instructions. Computer-readable storage media include, but are not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Computer-readable storage media may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, a non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the computer-readable storage medium, the various methods described above can be performed.

[0109] Example 5:

[0110] Fifthly, this application proposes a computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of the method described in any of the first aspects.

[0111] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0112] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0113] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0114] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0115] In the embodiments provided in this disclosure, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0116] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a controller to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0117] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein.

[0118] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing this application, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of this application.

Claims

1. A method for optimizing a supply chain network, characterized in that, include: Establish an initial supply chain network with defined rapid response capabilities for each node; The establishment of the initial supply chain network, which defines rapid response capabilities for each node, includes: Define the rapid response capabilities of each node; The initial load of each node is determined based on its fast response capability and fast response capability adjustment coefficient. The node capacity of each node is determined based on the initial load and the node's rapid response capability. The load redistribution rules of the initial supply chain network are determined based on the rapid response capabilities of each node. The establishment of the initial supply chain network, which defines rapid response capabilities for each node, also includes: The initial load of each node is determined by the following formula: 1+ ) in, Let be the initial load of node i. λ is the degree value of node i; x represents the load index; λ is the adjustment coefficient of the node's fast response capability, used to adjust the impact of fast response capability on the initial load; This represents the fast response capability of node i. The node capacity includes an upper capacity limit and a lower capacity limit, wherein the upper capacity limit is determined by the following formula: β represents the capacity limit, μ represents the capacity growth factor, and μ represents the adjustment coefficient of the rapid response capability on the capacity limit. The lower limit of capacity is determined by the following formula: in, Here, y represents the capacity lower limit, and ν represents the adjustment coefficient of the rapid response capability on the capacity lower limit. The load redistribution rules are as follows: in, This represents the load that was distributed from the failed node to its neighboring nodes; , where is the remaining capacity of neighbor node j. Mark the capacity limit of neighbor node j. This indicates the current load of neighbor node j; η Γi represents the adjustment coefficient of rapid response capability to load distribution, and Γi represents the set of neighboring nodes of the failed node i. This indicates the fast response capability of neighbor node j. Li This represents the load of the failed node i. To represent a constant, preventing errors such as multiplying by 0; After neighbor node j accepts the load redistributed by the failed node i, the load of neighbor node j is: In the formula, The current load of neighbor node J after it has accepted the redistributed load is given. , , This represents the load received by node j. This represents the load fluctuation caused by node j after it accepts the load. It is a factor that controls the amplitude of fluctuations; Calculate the overall centrality index of each node in the initial supply chain network; The initial supply chain network is simulated and analyzed based on the comprehensive centrality index, and the simulation analysis results are obtained. Based on the simulation analysis results, the optimized values ​​for each parameter in the initial supply chain network are determined; The initial supply chain network is optimized based on the optimization values ​​to obtain an optimized supply chain network.

2. The method according to claim 1, characterized in that, The step of performing simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtaining simulation analysis results includes: Nodes whose comprehensive centrality index meets the preset conditions are identified as target nodes; Perform a simulated attack on the target node; During the simulated attack, different values ​​were set for each parameter of the initial supply chain network; Calculate the cascading failure index of each parameter during the simulated attack process; Multiple cascade failure index-parameter data tables are generated based on the cascade failure indices described for each parameter.

3. The method according to claim 2, characterized in that, The step of determining the optimized values ​​for each parameter in the initial supply chain network based on the simulation analysis results includes: The optimized value is determined by the parameter value corresponding to the maximum cascading failure index in each of the cascading failure index-parameter data tables.

4. The method according to claim 1, characterized in that, The method further includes: The DNN model is combined with the LightGBM model to form an initial ensemble model; The initial ensemble model is optimized using the Nadam optimizer to form an optimized ensemble model. During the operation of the optimized supply chain network, the optimized integrated model is used to predict the nodes of the optimized supply chain network. Optimize the load on nodes at risk based on the prediction results for the nodes.

5. The method according to claim 1, characterized in that, The parameters that need to be optimized for the initial supply chain network include: the node's rapid response capability, the adjustment coefficient of the node's influence capability, the capacity growth factor, the adjustment coefficient of the capacity ceiling, the capacity floor parameter, the adjustment coefficient of the capacity floor, the adjustment coefficient of the rapid response capability on load distribution, and the factor for controlling the fluctuation range.

6. A supply chain network optimization device, characterized in that, include: The first execution module is used to establish an initial supply chain network that defines rapid response capabilities for each node; The establishment of the initial supply chain network, which defines rapid response capabilities for each node, includes: Define the rapid response capabilities of each node; The initial load of each node is determined based on its fast response capability and fast response capability adjustment coefficient. The node capacity of each node is determined based on the initial load and the node's rapid response capability. The load redistribution rules of the initial supply chain network are determined based on the rapid response capabilities of each node. The establishment of the initial supply chain network, which defines rapid response capabilities for each node, also includes: The initial load of each node is determined by the following formula: 1+ ) in, Let be the initial load of node i. λ is the degree value of node i; x represents the load index; λ is the adjustment coefficient of the node's fast response capability, used to adjust the impact of fast response capability on the initial load; This represents the fast response capability of node i. The node capacity includes an upper capacity limit and a lower capacity limit, wherein the upper capacity limit is determined by the following formula: β represents the capacity limit, μ represents the capacity growth factor, and μ represents the adjustment coefficient of the rapid response capability on the capacity limit. The lower limit of capacity is determined by the following formula: in, Here, y represents the capacity lower limit, and ν represents the adjustment coefficient of the rapid response capability on the capacity lower limit. The load redistribution rules are as follows: in, This represents the load that was distributed from the failed node to its neighboring nodes; , where is the remaining capacity of neighbor node j. Mark the capacity limit of neighbor node j. This indicates the current load of neighbor node j; η Γi represents the adjustment coefficient of rapid response capability to load distribution, and Γi represents the set of neighboring nodes of the failed node i. This indicates the fast response capability of neighbor node j. Li This represents the load of the failed node i. To represent a constant, preventing errors such as multiplying by 0; After neighbor node j accepts the load redistributed by the failed node i, the load of neighbor node j is: In the formula, The current load of neighbor node J after it has accepted the redistributed load is given. , , This represents the load received by node j. This represents the load fluctuation caused by node j after it accepts the load. It is a factor that controls the amplitude of fluctuations; Calculate the overall centrality index of each node in the initial supply chain network; The initial supply chain network is simulated and analyzed based on the comprehensive centrality index, and the simulation analysis results are obtained. Based on the simulation analysis results, the optimized values ​​for each parameter in the initial supply chain network are determined; The initial supply chain network is optimized based on the optimization values ​​to obtain an optimized supply chain network; The first calculation module is used to calculate the comprehensive centrality index of each node in the initial supply chain network; The second execution module is used to perform simulation analysis on the initial supply chain network based on the comprehensive centrality index and obtain the simulation analysis results; The first determining module is used to determine the optimized values ​​of each parameter in the initial supply chain network based on the simulation analysis results. The first optimization module is used to optimize the initial supply chain network based on the optimization values ​​to obtain an optimized supply chain network.

7. A computer electronic production equipment, characterized in that, It includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 5.

Images