A method for managing a network infrastructure system based on multi-dynamics coupling

Abstract

The application relates to the field of complex systems and network science, and provides a network-type infrastructure system management method based on multi-dynamics coupling, which comprises the following steps: modeling an infrastructure system existing in a preset region in a current period and subjected to disturbance to obtain a network to be processed; under the action of LLRM and GLRM, a plurality of multi-dynamics mechanism coupling models corresponding to the network to be processed are built; dynamics simulation is performed on each multi-dynamics mechanism coupling model to obtain a cascading failure characteristic simulation result under each multi-dynamics mechanism coupling model; according to an evaluation target of the infrastructure system in the current period, the cascading failure characteristic simulation result is used to determine a target coupling model of the infrastructure system from the plurality of multi-dynamics mechanism coupling models, and the infrastructure system is managed based on a dynamics mechanism in the target coupling model. The application can realize effective management of the infrastructure system existing in the current period and subjected to disturbance.

Description

Technical Field

[0001] This application relates to the field of complex systems and network science, and to, but is not limited to, a management method for network infrastructure systems based on multi-dynamic coupling. Background Technology

[0002] With the rapid advancement of science and technology and the high-speed development of society, critical infrastructure systems such as power systems, water supply systems, transportation systems, and communication systems have evolved into complex mega-systems with network characteristics. While these complex mega-systems efficiently safeguard people's daily lives, their fault propagation also exhibits highly nonlinear propagation characteristics. Complex networks are an important theory for studying complex mega-systems. By modeling the units and relationships of complex mega-systems as network nodes and edges, complex networks can characterize the topological structure of complex mega-systems. Based on the topological structure, combined with dynamic functions characterizing functional interactions, the dynamic characteristics of complex mega-systems can be studied.

[0003] In existing technologies, cascade dynamics describes how an initial disturbance in a complex giant system is gradually amplified through interactions between nodes or edges, leading to a series of successive failures. However, most studies focus on the influence of a single dynamic mechanism, rarely considering the coupling effects of multiple coexisting dynamic mechanisms. In reality, the failure propagation process of complex giant systems may be simultaneously influenced by multiple dynamic mechanisms, and their coupling effects cannot be decoupled for analysis; that is, the results of the combined effects of multiple dynamic mechanisms cannot be predicted by analyzing the conclusions of a single dynamic study. Therefore, there is an urgent need to provide a cascade dynamics modeling and analysis scheme capable of characterizing the coupling effects of multiple dynamic mechanisms. Summary of the Invention

[0004] Based on the above problems, this application provides a management method for network infrastructure systems based on multi-dynamic coupling. It aims to provide a technical framework for infrastructure systems with disturbances, from mechanism modeling to simulation optimization and strategy implementation, so as to achieve comprehensive performance evaluation of infrastructure systems with disturbances in the current period, thereby enabling effective management of infrastructure systems with disturbances in the current period.

[0005] The technical solution of this application embodiment is implemented as follows: This application provides a management method for a network infrastructure system based on multi-dynamic coupling. The method includes: modeling an infrastructure system with disturbances in a preset area during the current time period to obtain a network to be processed; constructing multiple multi-dynamic mechanism coupling models corresponding to the network to be processed under the combined action of Local Load Redistribution Mechanism (LLRM) and Global Load Redistribution Mechanism (GLRM); performing dynamic simulation on each of the multiple multi-dynamic mechanism coupling models to obtain the simulation results of cascading failure characteristics under each multi-dynamic mechanism coupling model; determining the target coupling model of the infrastructure system from the multiple multi-dynamic mechanism coupling models based on the cascading failure characteristic simulation results, according to the evaluation target of the infrastructure system during the current time period, and based on the dynamic mechanisms within the target coupling model, managing the infrastructure system.

[0006] In some embodiments, modeling an infrastructure system with disturbances in a preset area during the current time period to obtain a network to be processed with cascading failures includes: abstracting the units and relationships operating within the infrastructure system into network nodes and connecting edges, respectively; and performing state labeling and relationship quantification on the network nodes and connecting edges to obtain the network to be processed.

[0007] In some embodiments, the multi-dynamic mechanism coupling model includes: a two-type load-two-dynamic-mechanism independent propagation model (TTIP); and, under the combined action of the local load redistribution mechanism (LLRM) and the global load redistribution mechanism (GLRM), constructing a multi-dynamic mechanism coupling model corresponding to the network to be processed, including: obtaining faulty nodes in the network to be processed. Transferred to node under LLRM action Local redistribution of load and based on local load redistribution For nodes Adjust the current load to obtain the node New local load Based on the faulty nodes of the network to be processed Global load balancing performed under GLRM results in node... New global load ;exist or Representation Nodes In the event of overload failure, a TTIP (Time-to-Incident Propagation Protocol) is constructed to cascade fault propagation based on LLRM (Limited Load Management) for local loads and GLRM (Global Load Management). and They are nodes The total initial capacity under LLRM and GLRM.

[0008] In some embodiments, the multi-dynamic mechanism coupling model includes: a two-load-two-dynamic-mechanism co-operation model (TTWT); and a multi-dynamic mechanism coupling model corresponding to the network to be processed, constructed under the combined action of the local load redistribution mechanism (LLRM) and the global load redistribution mechanism (GLRM), including: obtaining faulty nodes in the network to be processed. Under the action of LLRM, the transfer to the node Local redistribution of load and based on local load redistribution For nodes Adjust the current load to obtain the node New local load Based on the faulty nodes of the network to be processed Global load balancing performed under GLRM results in node... New global load ;exist Representation Nodes The system constructs TTWTs corresponding to cascading fault propagation based on LLRM for local load and GLRM for global load; among which, For nodes The initial total capacity set under the combined action of LLRM and GLRM, and , and All of these are adjustable capacity parameters.

[0009] In some embodiments, the multi-dynamic mechanism coupling model includes: a load-two-dynamic-mechanism coupling effect model (OTCE); and a multi-dynamic mechanism coupling model corresponding to the network to be processed, constructed under the combined action of the local load redistribution mechanism (LLRM) and the global load redistribution mechanism (GLRM), including: integrating faulty nodes in the network to be processed. Transferred to node under LLRM action Local redistribution of load And nodes after global load balancing based on GLRM for other non-failed nodes in the network to be processed. Background load: , obtain node New current load: ;in, The number of nodes in the network to be processed; For nodes As a receiving unit, it receives data from other non-failed [receivers / receivers]. The load stream received by each node; For all node pairs in the network to be processed The load passes through the nodes The sum of the loads; in Representation Nodes In the event of overload failure, an OTCE is constructed to propagate cascading faults based on LLRM for local loads and GLRM for global loads; among which, For nodes The initial total capacity set under the combined coupling effect of LLRM and GLRM.

[0010] In some embodiments, an internally operating unit is detached from the main body of the infrastructure system; and / or an internally operating unit experiences an overload failure, thus identifying the infrastructure system as having a disturbance.

[0011] In some embodiments, dynamic simulation is performed on each of the multiple multi-dynamic mechanism coupling models to obtain the simulation results of cascading failure characteristics under each multi-dynamic mechanism coupling model. This includes: for each multi-dynamic mechanism coupling model, performing coupling relationship mapping and failure threshold calibration on the initial topology and initial load distribution of the multi-dynamic mechanism coupling model to obtain the simulation initialization configuration file of the multi-dynamic mechanism coupling model; performing cascading propagation iteration within discrete time steps on the simulation initialization configuration file, and recording the load over-limit node and failure propagation range at each time step to obtain the cascading failure characteristic simulation results of each multi-dynamic mechanism coupling model, including failure timing and cascading scale.

[0012] The beneficial effects of the technical solutions provided in this application include at least the following: The management method for network infrastructure systems based on multi-dynamic coupling provided in this application involves the following steps: First, a model is created for the infrastructure system in a preset area experiencing disturbances during the current time period, resulting in a network to be processed. Then, under the combined action of LLRM and GLRM, multiple multi-dynamic mechanism coupling models corresponding to the network to be processed are constructed. Dynamic simulations are performed on each of the multiple multi-dynamic mechanism coupling models to obtain simulation results of cascading failure characteristics under each model. Finally, based on the evaluation objectives of the infrastructure system during the current time period and the simulation results of cascading failure characteristics, a target coupling model for the infrastructure system is determined from the multiple multi-dynamic mechanism coupling models. The infrastructure system is then managed based on the dynamic mechanisms within the target coupling model. Thus, on the one hand, modeling infrastructure systems with disturbances can accurately depict the initial distribution characteristics of disturbances on the topology and functional nodes of the infrastructure system, providing a high-fidelity network foundation for subsequent coupling analysis. Furthermore, by introducing LLRM and GLRM to jointly construct multiple multi-dynamic mechanism coupling models, it is possible to comprehensively reflect the dual dynamic process of local failure propagation and global load coordination adjustment in the infrastructure system, thereby avoiding the limitation of a single mechanism failing to adequately describe the actual failure propagation process. On the other hand, the simulation results of cascading failure characteristics under multiple multi-dynamic mechanism coupling models obtained from dynamic simulation can be used as a comparison basis, and the optimal coupling model can be selected based on specific evaluation objectives to guide management decisions. In this way, a technical framework from mechanism modeling to simulation optimization to strategy implementation can be provided for infrastructure systems with disturbances, so as to achieve comprehensive performance evaluation of infrastructure systems with disturbances in the current period, thereby enabling effective management of infrastructure systems with disturbances in the current period.

[0013] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the technical solutions provided in the embodiments of this application. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 A flowchart illustrating a management method for a network infrastructure system based on multi-dynamic coupling, provided as an embodiment of this application; Figure 2The following are schematic diagrams of simulation results for the TTIP corresponding to the general generative network model provided in this application embodiment under different combinations of redistribution parameters. Part (a) in the figure is a schematic diagram of the scale of failed nodes under LLRM, part (b) in the figure is a schematic diagram of the scale of failed nodes under GLRM, and part (c) in the figure is a schematic diagram of the total scale of network failed nodes. Figure 3 This is a schematic diagram illustrating the comparative analysis of the general generative network model based on TTIP across various cascade failure node scales provided in the embodiments of this application; portions (a1) to (a3) ​​in the diagram represent different β values. L Below, with β G A schematic diagram showing the relationship between different failure node sizes. Parts (b1) to (b3) in the diagram represent different β values. G Below, with β L A schematic diagram showing the corresponding relationships between different types of failure node sizes; Figure 4 The following is a schematic diagram of the simulation results of TTIP corresponding to the route network in the xx area provided in the embodiments of this application under different combinations of redistribution parameters; part (a) in the figure is a schematic diagram of the scale of failed nodes under LLRM, part (b) in the figure is a schematic diagram of the scale of failed nodes under GLRM, and part (c) in the figure is a schematic diagram of the total scale of network failed nodes. Figure 5 The following is a schematic diagram of the cascading failure simulation results of TTWT on a general generated network model and a route network within an xx region, provided for the embodiments of this application. Part (a) in the figure is a schematic diagram of the cascading failure simulation results on the general generated network model, and part (b) in the figure is a schematic diagram of the cascading failure simulation results on a route network within an xx region. Figure 6 The figure shows the cascading failure simulation results of OTCE in the general generated network model and the route network in the xx region, which are provided in the embodiments of this application. Part (a) of the figure shows the cascading failure simulation results in the general generated network model, and part (b) of the figure shows the cascading failure simulation results in the route network in the xx region. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. The following embodiments are used to illustrate this application, but are not intended to limit the scope of this application. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0016] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0017] It should be noted that the terms "first, second, and third" used in the embodiments of this application are merely to distinguish similar objects and do not represent a specific order of objects. It is understood that "first, second, and third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0018] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments of this application pertain. It should also be understood that terms such as those defined in general dictionaries should be understood to have a meaning consistent with their meaning in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0019] In existing technologies, major / catastrophic accidents in infrastructure systems, such as the snow disaster in a certain region in 2008 that caused multiple power grid disconnections and affected more than a dozen areas, and the massive traffic jam in a certain international city in 2014 with a total length of 344 kilometers, are inseparable from the impact of cascading failures.

[0020] The research on numerous cascade dynamics-related models mainly includes the following: 1. The global cascading effect of random networks under binary models reveals that heterogeneity plays a dual and ambiguous role in determining system stability: the stronger the heterogeneity of node failure threshold, the more susceptible the system is to global cascading effects; the stronger the heterogeneity of network degree distribution, the more robust the network is and the less vulnerable it is.

[0021] 2. Using a sandpile model to study the load cascading of spatial networks, we found that spatial factors promote the occurrence of medium-sized avalanches and inhibit large-scale avalanche outbreaks. We also explained this differentiated evolution mechanism based on triangular topology.

[0022] 3. The coupled image lattice model was used to analyze cascading failures on scale-free networks. The results show that for highly heterogeneous networks, deliberate attacks on highly important nodes are more likely to cause large-scale cascading failures than random attacks.

[0023] 4. By introducing the fiber bundle model into scale-free networks, network nodes are defined to have the same initial load, and the node capacity is set to meet a certain statistical distribution. After a node fails, its load will be evenly distributed to its neighboring nodes. The study found that the higher the uniformity of the node capacity distribution, the stronger the network's robustness against cascading failures.

[0024] 5. Assign an initial load range [L] to each unit. min L max Random values ​​are set, and a failure threshold L is set. fail The study found that under low load, the distribution of the number of failed components exhibits a near-exponential truncated tail characteristic, and the risk of large-scale cascading failures is low. However, the system has a critical load threshold. Exceeding this threshold will cause the distribution of the number of failed components to have a power-law region, and the gradient of the average number of failed components will increase sharply. As the load increases above the critical load, the distribution of the number of failed components saturates, and there is a greater risk of cascading failures.

[0025] 6. A global load capacity model is proposed, which defines node load based on network traffic formed by the flow of energy between nodes along the shortest path. Changes in network structure lead to changes in the shortest path between nodes, resulting in a global redistribution of network load. When the node load exceeds its capacity, a cascading failure occurs.

[0026] 7. The initial load of nodes is defined as a degree centrality correlation function. This definition is not only superior to the load averaging setting, but also overcomes the requirement of global network information for load calculation in the models proposed in related literature. It greatly reduces the computational complexity in large-scale networks. At the same time, adjustable parameters are set to select appropriate load functions for different network structures. The redistribution rule adopts an optimal allocation strategy based on local topology information, which is consistent with the flow logic of real-world energy redistribution.

[0027] 8. Consider the functional dependencies between network nodes and establish dependency edges to characterize the functional dependency constraints between nodes: if any node at either end of a dependency edge fails, the other node will also fail synchronously. This model can be extended to analyze multi-layer networks and is widely used in the study of physical information systems.

[0028] 9. Introduce dependency relationships into a single-layer network topology and analyze how a certain proportion of dependency edges in the network affect the percolation effect of the network.

[0029] However, most existing studies only focus on the influence of a single dynamic mechanism, and rarely consider the coupling effect of multiple dynamic mechanisms coexisting. In reality, the failure propagation process of complex giant systems may be simultaneously affected by multiple dynamic mechanisms, and the coupling effect between them cannot be decoupled and analyzed. That is, it is impossible to predict the result of the combined effect of multiple dynamic mechanisms by analyzing the conclusions of a single dynamic study.

[0030] In other words, existing research on cascade dynamics largely focuses on the fault propagation laws under a single dynamic mechanism, making it difficult to effectively explain the nonlinear coupling effects generated when multiple dynamic mechanisms coexist. Since disturbance propagation in real complex giant systems is often driven by multiple heterogeneous dynamic processes, and these processes cannot be decoupled for analysis, it is impossible to directly predict the overall behavior of the system under the coupling of multiple mechanisms based on the conclusions of single dynamic studies. This limitation restricts the applicability and accuracy of existing methods in vulnerability assessment and fault evolution prediction for real complex giant systems. Based on this, this application constructs three dynamic coupling modes according to the dynamic mechanisms of two typical load-capacity models to study the cascade process under the combined action of multiple dynamic mechanisms. This aims to explore the impact of cascade failure under different coupling modes of multiple dynamic mechanisms, and thus determine management methods for real complex giant systems based on the simulation results of cascade failure under different coupling modes and the corresponding assessment objectives.

[0031] Example 1: refer to Figure 1 This is a flowchart illustrating a management method for a network infrastructure system based on multi-dynamic coupling, provided in an embodiment of this application. Figure 1 The following explanation is provided: Step 101: Model the infrastructure system in the preset area that is disturbed in the current time period to obtain the network to be processed.

[0032] In some embodiments, the infrastructure system may refer to the power grid, transportation network, water supply network, communication network, etc., within a preset area. Meanwhile, the current time period may refer to the current hour, the current day, etc., and this application does not impose any limitations on it.

[0033] In some embodiments, an internally operating unit is detached from the main body of the infrastructure system; and / or an internally operating unit experiences an overload failure, thus identifying the infrastructure system as having a disturbance.

[0034] In some embodiments, when certain units operating within an infrastructure system detach from their main support structure or lose their normal collaborative relationship with the main system, and / or when certain units operating within an infrastructure system experience overload faults such as operating load exceeding their design threshold, resulting in decreased processing capacity or functional abnormalities, the infrastructure system is considered to be in a disturbed state, i.e., there is a disturbance.

[0035] Here, for infrastructure systems within a preset area that are subject to disturbances during the current time period, their local autonomous behavior or failure behavior may deviate from the overall control objective, thereby affecting the stability, continuity, or security of the external services provided by the infrastructure system within the preset area.

[0036] For example, consider a power grid in a certain region: if one of the distributed energy storage devices switches to islanded operation mode due to a communication interruption and is no longer subject to unified control by the dispatch center (i.e., disconnected from the main system); simultaneously, a nearby transmission line is under prolonged overload due to a sudden increase in load, and the equipment temperature continues to exceed the standard (i.e., overload fault). In this case, the power grid is immediately identified as an infrastructure system with disturbances, which may lead to a decline in power quality or the risk of local power outages.

[0037] In some embodiments, an infrastructure system experiencing disturbances within a preset region during the current time period is modeled to obtain a network to be processed that has experienced cascading failures; that is, the network to be processed can characterize the infrastructure system. Each component unit and A relationship is typically represented by an undirected, unweighted graph. This will be described. The set of nodes (component units) is included. It can be represented as: The set of edges (associations) It can be represented as Here, adjacency matrices can also be used. Represent the network to be processed; where, if the node and nodes If there are connecting edges between them, then ,otherwise , and .

[0038] Correspondingly, step 101 above can be implemented by following steps 1011 and 1012. Figure 1 (Not yet realized in China) Step 1011: Abstract the units and relationships operating within the infrastructure system into network nodes and connection edges, respectively.

[0039] Step 1012: Perform state labeling and relation quantification on network nodes and connecting edges to obtain the network to be processed.

[0040] In some embodiments, firstly, the units operating within the infrastructure system (which can be physical units or functional units, specifically power plants, base stations, and pumping stations) are abstracted as network nodes, and the relationships between units (such as supply relationships, control relationships, or dependency relationships) are abstracted as connection edges; then, each network node and connection edge is calibrated in terms of state (such as normal, overload, or failure) and its relationships are quantified (such as capacity, current load, and fault propagation probability), thereby forming a computable directed or undirected weighted network, i.e., the network to be processed.

[0041] Here, taking a transportation system as an example of an infrastructure system, we first abstract intersections and roads in the transportation system as network nodes and connecting edges, respectively. Then, we define traffic flow as load and the load processing capacity of intersections as capacity, to achieve network state labeling and relationship quantification of the transportation system. When a traffic accident causes a decrease in traffic flow capacity (capacity reduction) at the accident point, on the one hand, the accident increases the traffic load at the accident point, causing congestion (overload) at the intersection, which is marked as overloaded. Continuous congestion will cause the traffic load to spread to adjacent intersections, leading to traffic overload at adjacent intersections as well, resulting in a local cascading diffusion phenomenon. On the other hand, traffic in the transportation system network will change its travel routes to avoid the accident congestion point, causing a redistribution of network load, resulting in increased traffic flow on some roads, exceeding their capacity (overload), resulting in a global cascading diffusion phenomenon. By quantifying the relationship between traffic flow and capacity, and the dynamic change rules of traffic flow, we construct a network to be processed that can be used to analyze the cascading diffusion process of the transportation system.

[0042] In this way, the difficult-to-observe and nonlinear cascading failure process within the infrastructure system can be transformed into a structured, quantifiable, and iteratively computed network model, thereby providing parameter support for subsequent simulation of cascading failure characteristics.

[0043] Step 102: Under the combined action of the local load redistribution mechanism LLRM and the global load redistribution mechanism GLRM, a coupled model of multiple dynamic mechanisms corresponding to the network to be processed is built.

[0044] In some embodiments, cascading failure is an important form of network fault propagation, and the load-capacity model is a widely studied dynamic model. This application describes two of the most widely used load-capacity cascading failure models and their dynamic mechanisms: (1) Local Load Rebalancing Model and Local Load Rebalancing Mechanism (LLRM): In LLRM, node load is defined as a correlation function based on node degree centrality, where nodes in the network to be processed... Local load Represented as: Formula (1); in, Represents a node Degree centrality metric, i.e., the degree of connection between nodes in a network. The number of directly connected neighbor nodes; the load adjustability parameters are satisfied. and .

[0045] Here, the degree centrality index of nodes in the network can be calculated based on the adjacency matrix. : Formula (2); in, These are elements in the adjacency matrix, as mentioned above, if a node... and nodes If there are connecting edges between them, then ,otherwise .

[0046] It should be noted that the initial capacity of a node refers to the capacity of that node. The maximum load that can be withstood is usually defined as a linear or nonlinear function of the load. In this application, the nodes can be set based on the following formula (3). initial capacity The nonlinear functional relationship of the load: Formula (3); in, and These are the adjustable parameters for the LLRM capacity. Represents a node Redundancy capacity, i.e., the node The amount of redistributed load that can be tolerated, i.e., the amount of local load it can bear. In addition, it can also handle the additional load redistributed from other nodes.

[0047] It should be noted that, in actual operation, nodes initial capacity Typically, nodes are fixed. However, in specific situations such as accidents in transportation networks or overheating of equipment in power grids, nodes... The current available capacity will be based on the node. initial capacity The dynamic reduction yields the node's dynamic value during the load redistribution process. The actual maximum capacity that can be sustained. In this application, it is measured in terms of nodes. initial capacity Let's take a fixed and unchanging example as an example.

[0048] Here, LLRM refers to the load redistribution process based on node local topology information. When a node... In the event of a failure, the load will simply be transferred to the pending network and nodes. Directly connected neighbor nodes. The specific behavior of LLRM mentioned in existing literature is: when a node... In the event of a failure, the load is redistributed to other adjacent nodes according to the optimal allocation rule shown in formula (4). This optimal allocation rule ensures that the redistributed load is shifted more towards nodes with larger loads. Among these, the node... Transfer to adjacent nodes The redistribution of load is .

[0049] Formula (4); in, For nodes in the network to be processed Local load, For nodes in the network to be processed The set of adjacent nodes; To process nodes in the network All neighboring nodes within the set of adjacent nodes Local load Perform summation.

[0050] Here, if node If the total received load satisfies the relationship shown in formula (5), then the node An overload failure will trigger a new round of load redistribution. Assuming the load redistribution impact is a transient process, if a node... If the impact of the current load redistribution is successfully withstood, it is believed that the continued impact of the load redistribution can be eliminated by disconnecting the link. Based on this, this application only considers the superposition of loads redistributed in the same period, and does not consider the cumulative effect of load redistribution in different periods.

[0051] Formula (5); in, To start from the node where the failure occurred Transfer to its neighboring node The load; For nodes in the network to be processed The set of adjacent nodes; Indicates a node All neighboring nodes Perform summation; This refers to the set of faulty nodes in the network to be processed. For nodes Redundancy capacity.

[0052] (2) Global Load Rebalancing Model and Global Load Rebalancing Mechanism (GLRM): The global load redistribution model, proposed by Motter et al., defines the load on nodes and its redistribution rules based on the shortest path principle of energy flow between nodes in a network. It assumes that every pair of nodes in the network sends one unit of energy to each other at any given time, with the energy flow based on the shortest path in the network. Represents ordered node pairs Energy flow between nodes The contribution of ordered nodes. The shortest path between them passes through the nodes. ,remember ;otherwise If multiple shortest paths exist between a pair of nodes, the energy flow is evenly distributed. Therefore, a node is defined. The load is: Formula (6); At the same time, define nodes capacity for: Formula (7); in, and This is the adjustable capacity parameter of the GLRM. For nodes Redundancy capacity.

[0053] The GLRM process is as follows: If a node fails, the node and its edges are removed. This changes the shortest paths between some nodes on the network, altering the energy flow paths between the remaining nodes and causing a global redistribution of network load. If a node... Emerging under new loads Then the node If a node fails due to overload, the newly failed node and its connected edges are removed, which in turn causes a change in the energy flow path between the remaining nodes in the network, resulting in a redistribution of network load and the propagation of cascading faults until no new failed nodes appear in the network.

[0054] In some embodiments, the multiple multi-dynamic mechanism coupling model includes: a two-type loads-two-dynamic-mechanisms independent propagation model (TTIP); correspondingly, step 102 above can be implemented by the following steps 1021 to 1023. Figure 1 (not shown in the image) Step 1021: Obtain the faulty nodes in the network to be processed. Transferred to node under LLRM action Local redistribution of load Δ and based on local load redistribution For nodes Adjust the current load to obtain the node New local load .

[0055] Step 1022: Based on the faulty nodes of the network to be processed Global load balancing performed under GLRM results in node... New global load .

[0056] Step 1023, in or Representation Nodes In the event of overload failure, a TTIP is constructed to cascade the fault propagation based on the LLRM for local load and the GLRM for global load.

[0057] in, and They are nodes The total initial capacity under LLRM and GLRM.

[0058] In some embodiments, TTIP can be used to characterize two types of loads with different properties existing simultaneously in the network to be processed, which cascade and spread under their respective independent dynamic mechanisms. Faulty node Failure for any reason will result in two types of load: new local load. (i.e., fault node) Transferred to node under LLRM action Local redistribution of load With nodes (sum of current loads) and new global load Simultaneously, cascading faults propagate according to their respective dynamic mechanisms. During the propagation of cascading faults, the effects of the two dynamic mechanisms are independent, but any load exceeding its corresponding capacity will cause the node to fail. Invalid.

[0059] In some embodiments, the multiple dynamic mechanism coupling model includes: a two types of loads two dynamic mechanisms working together model (TTWT); correspondingly, step 102 above can be implemented by steps 1024 to 1026. Figure 1 (not shown in the image) Step 1024: Obtain the faulty nodes in the network to be processed. Transferred to node under LLRM action Local redistribution of load and based on local load redistribution For nodes Adjust the current load to obtain the node New local load .

[0060] Step 1025: Based on the faulty nodes of the network to be processed Global load balancing performed under GLRM results in node... New global load .

[0061] Step 1026, in Representation Nodes In the event of overload failure, the corresponding TTWT is constructed for cascading fault propagation based on LLRM for local load and GLRM for global load.

[0062] in, For nodes The initial total capacity set under the combined action of LLRM and GLRM, and , and All of these are adjustable capacity parameters.

[0063] In some embodiments, TTWT can be used to represent that each node in a network bears two loads with the same properties but different dynamic mechanisms, and the total load of the nodes is the sum of these two loads (node ​​loads). New local load ,node New global load The load is superimposed on the total load, and the node capacity is defined based on the total load. During cascading failure, these two types of loads are redistributed under their respective dynamic mechanisms. Whether a node experiences cascading failure depends on whether the node's total load (the sum of the new local load and the new global load) exceeds its capacity. TTWT can characterize the combined effect of two loads with the same properties on the network.

[0064] In some embodiments, the multiple multi-dynamic mechanism coupling model includes: a one-type-load two dynamic mechanisms coupling effect model (OTCE); correspondingly, step 102 above can be implemented by the following steps 1027 and 1028. Figure 1(not shown in the image) Step 1027: Integrate faulty nodes in the network to be processed. Transferred to node under LLRM action Local redistribution of load And nodes after global load balancing based on GLRM for other non-failed nodes in the network to be processed. Background load: , obtain node New current load: .

[0065] in, The number of nodes in the network to be processed; For nodes As a receiving unit, from other non-failed The load stream received by each node; For all node pairs in the network to be processed The load passes through the nodes The sum of the loads.

[0066] Step 1028, in Representation Nodes In the event of overload failure, the corresponding OTCE is constructed by cascading fault propagation based on LLRM for local load and GLRM for global load.

[0067] in, For nodes The initial total capacity set under the combined coupling effect of LLRM and GLRM.

[0068] In some embodiments, OTCE is based on the global load redistribution model proposed by Motter et al., with the initial load defined as the faulty node. Under the coupling of LLRM and GLRM, the transfer to the node The load is superimposed. When a node fails, on the one hand, based on LLRM, the load currently borne by that node is redistributed to its neighboring nodes; on the other hand, based on GLRM, changes in the shortest path of energy flow between each pair of nodes lead to network load reconfiguration. The new load of a node is a superposition of the locally redistributed load and the load after the full network reconfiguration.

[0069] In this way, under the combined effect of LLRM and GLRM, the coupled multi-dynamic mechanism model corresponding to the network to be processed can simultaneously characterize the rapid, local load migration phenomenon triggered by overload within nodes or local subnets, as well as the slow time-scale load rebalancing phenomenon across regions based on global topology and state information, under different operating mechanisms. This enables a cross-scale, nonlinear dynamic description of the congestion propagation, self-healing recovery, and phase transition behavior of the network to be processed.

[0070] Step 103: Perform dynamic simulation on each of the multiple dynamic mechanism coupling models to obtain the simulation results of the cascading failure characteristics under each multiple dynamic mechanism coupling model.

[0071] In some embodiments, Matlab can be used to perform dynamic simulations on each multi-dynamic mechanism coupled model. When performing dynamic simulations on each multi-dynamic mechanism coupled model, the unique load allocation mechanism within that model must be analyzed. This mechanism defines how, after an initial disturbance, the load it carries (e.g., current, information flow, traffic flow, material flow) is redistributed to the remaining healthy units in the infrastructure system. Different load type relationships, allocation logics, and capacity performance directly lead to significant differences in the spatiotemporal scale, propagation path, and final collapse mode of cascading failures. By quantitatively simulating these coupled dynamics, the vulnerability and resilience characteristics of the infrastructure system under specific failure scenarios, such as local disturbances, can be revealed.

[0072] In some embodiments, step 103 above can be implemented by steps 1031 and 1032. Figure 1 (not shown in the image) Step 1031: For each multi-dynamic mechanism coupling model, perform coupling relationship mapping and failure threshold calibration on the initial topology and initial load distribution of the multi-dynamic mechanism coupling model to obtain the simulation initialization configuration file of the multi-dynamic mechanism coupling model.

[0073] Step 1032: Perform cascading propagation iteration within discrete time steps on the simulation initialization configuration file, and record the load overload node and failure propagation range at each time step to obtain the cascading failure characteristic simulation results of each of the multi-dynamic mechanism coupling models, including the failure timing and cascading scale.

[0074] In some embodiments, for each multi-dynamic mechanism coupled model, dynamic coupling relationship mapping (e.g., in a traffic network, the relationship between global load change and local load change is: independent action, joint action, coupled action) and failure threshold calibration (e.g., setting the capacity upper limit, attenuation degree, capacity properties, etc. of each node) are performed to generate the simulation initialization configuration file of the multi-dynamic mechanism coupled model.

[0075] Here, the simulation initialization configuration file of the multi-dynamic mechanism coupling model must explicitly define two types of core rules: node failure principle (the node failure judgment principle of each of the three multi-dynamic mechanism coupling models has been given above) and internal load distribution mechanism (local load redistribution mechanism and global load redistribution, whether they are executed independently, act together, or are coupled).

[0076] Correspondingly, the simulation initialization configuration file is subjected to cascading propagation iteration within discrete time steps. Within each discrete time step, the following steps are executed sequentially: the internal load allocation mechanism completes the redistribution of failed loads, and based on the node failure principle, it determines the load-over-limit nodes and marks them as failed, updates the network topology and load distribution, and records the load-over-limit nodes and the failure propagation range in each discrete time step. At the same time, it records the triggering process corresponding to the failure principle and allocation mechanism adopted in each discrete time step, and obtains the cascading failure characteristic simulation results of each multi-dynamic mechanism coupled model, including the failure sequence and cascading scale.

[0077] Step 104: Based on the assessment objectives of the infrastructure system in the current period, and based on the simulation results of cascading failure characteristics, determine the target coupling model of the infrastructure system from multiple multi-dynamic mechanism coupling models, and manage the infrastructure system based on the dynamic mechanisms within the target coupling model.

[0078] In some embodiments, multiple multi-dynamic mechanism coupling models include: TTIP, TTWT, and OTCE. Correspondingly, firstly, based on the assessment objectives of the infrastructure system in the current period (e.g., minimizing the scale of cascading failures, maximizing system resilience, or ensuring critical load connectivity), and combined with simulation results based on cascading failure characteristics (e.g., load redistribution threshold, node redundancy capacity, failure propagation speed), the target coupling model that best reflects the risk characteristics of the infrastructure system in the current period is determined from TTIP, TTWT, and OTCE by comparing the fitting accuracy and predictive ability of different models for simulated failure events. Subsequently, based on the key dynamic mechanisms identified within the target coupling model (e.g., LLRM and GLRM, whether they are executed independently, act together, or are coupled), targeted system management strategies are formulated to ensure that the infrastructure system maintains an acceptable operating state after being disturbed.

[0079] For example, the infrastructure system is a power grid system in a certain region. If, during the current summer peak electricity consumption period, the evaluation objective of this power grid system is to prevent large-scale power outages caused by a single substation failure, then steps 101 to 103 above are performed on the power grid system to obtain the simulation results of the cascading failure characteristics under TTIP, TTWT, and OTCE associated with the power grid system. Here, the cascading failure of this power grid system mainly follows the load-temperature-capacity three-field coupled dynamic mechanism (i.e., the load increase leads to the equipment temperature increase, and the high temperature further reduces the equipment capacity, forming a positive feedback). Through the simulation results of the cascading failure characteristics under TTIP, TTWT, and OTCE, OTCE is determined as the target coupling model. Accordingly, the power grid system takes corresponding management measures for the thermodynamic mechanism within OTCE, such as: coordinated control measures: before the node load approaches its critical threshold, the air cooling system of the adjacent substation is actively started to reduce the equipment temperature, thereby dynamically increasing its remaining capacity; at the same time, some loads are smoothly transferred to redundant lines through the power flow controller. The above measures work together to prevent global load redistribution from being triggered by the overheating and shutdown of a single node, thereby effectively curbing the spread of cascading failures.

[0080] For example, the evaluation objectives include: operational efficiency objectives, risk resistance objectives, and recovery cost control objectives. Correspondingly, the step 104 above, "based on the evaluation objectives of the infrastructure system in the current period, and based on the simulation results of cascading failure characteristics, determine the target coupling model of the infrastructure system from multiple multi-dynamic mechanism coupling models," can be achieved through the following process: First, perform multi-objective quantification and weight allocation on the operational efficiency objectives, risk resistance objectives, and recovery cost control objectives of the infrastructure system in the current period to obtain a comprehensive decision vector. Then, based on the simulation results of cascading failure characteristics, perform normalized weighted comprehensive evaluation on each multi-dynamic mechanism coupling model in terms of operational efficiency, risk resistance, and recovery cost control to obtain a comprehensive simulation evaluation value for each multi-dynamic mechanism coupling model. Finally, determine the multi-dynamic mechanism coupling models whose cascading failure characteristic simulation values ​​are greater than or equal to the threshold range corresponding to the comprehensive decision vector as the target coupling models of the infrastructure system in the current period.

[0081] The management method for network infrastructure systems based on multi-dynamic coupling provided in this application involves the following steps: First, a model is created for the infrastructure system in a preset area experiencing disturbances during the current time period, resulting in a network to be processed. Then, under the combined action of LLRM and GLRM, multiple multi-dynamic mechanism coupling models corresponding to the network to be processed are constructed. Dynamic simulations are performed on each of the multiple multi-dynamic mechanism coupling models to obtain simulation results of cascading failure characteristics under each model. Finally, based on the evaluation objectives of the infrastructure system during the current time period and the simulation results of cascading failure characteristics, a target coupling model for the infrastructure system is determined from the multiple multi-dynamic mechanism coupling models. The infrastructure system is then managed based on the dynamic mechanisms within the target coupling model. Thus, on the one hand, modeling infrastructure systems with disturbances can accurately depict the initial distribution characteristics of disturbances on the topology and functional nodes of the infrastructure system, providing a high-fidelity network foundation for subsequent coupling analysis. Furthermore, by introducing LLRM and GLRM to jointly construct multiple multi-dynamic mechanism coupling models, it is possible to comprehensively reflect the dual dynamic process of local failure propagation and global load coordination adjustment in the infrastructure system, thereby avoiding the limitation of a single mechanism failing to adequately describe the actual failure propagation process. On the other hand, the simulation results of cascading failure characteristics under multiple multi-dynamic mechanism coupling models obtained from dynamic simulation can be used as a comparison basis, and the optimal coupling model can be selected based on specific evaluation objectives to guide management decisions. In this way, a technical framework from mechanism modeling to simulation optimization to strategy implementation can be provided for infrastructure systems with disturbances, so as to achieve comprehensive performance evaluation of infrastructure systems with disturbances in the current period, thereby enabling effective management of infrastructure systems with disturbances in the current period.

[0082] The management method for the network infrastructure system based on multi-dynamic coupling described above will be explained below with reference to a specific embodiment. However, it should be noted that this specific embodiment is only for better illustration of this application and does not constitute an improper limitation of this application.

[0083] When a network suffers initial damage, the fault will propagate through a cascading effect. The size of network mega-components is often used to quantify the robustness of the network. This application conducts simulation analysis on a general generative network model and a real network model: a flight route network within a certain region. The general generative network model is selected as a scale-free network model, following the construction method proposed by Albert and Barabási in their relevant documents. The network size is: =500, average degree satisfy Among them, average degree This characterizes the average number of nodes in the network to approximately [number missing]. There are connected edges, that is, A neighbor, That is, a node The number of connections; the route network within the xx region has 332 nodes and 2126 edges. Set the load adjustable parameter in the model corresponding to LLRM to: , To ensure that the total load of the models corresponding to LLRM and GLRM is consistent, the initial failure node is selected as the node with the largest sum of local and global loads. This will maximize the simultaneous occurrence of load redistribution fluctuations under both LLRM and GLRM, resulting in the maximum extent of network cascading failure propagation. To eliminate the influence of randomness, the simulation data of the generated network is the average of calculations on 40 independent networks.

[0084] Part 1: Cascade Failure Characteristics Analysis of TTIP like Figure 2 The diagram illustrates the simulation results of TTIP corresponding to the general generative network model under different redistribution parameters; the redistribution parameters (capacity parameters) include: local redistribution parameter β. L and global reallocation parameter β G Here, β L , or β G It can be equivalently considered that the cascading failure of the general generative network model is only affected by a single LLRM or GLRM, which can be used to compare and analyze the differences in the effects of single dynamic mechanism and multi-dynamic coupling cascade.

[0085] See Figure 2 Part (a) shows that when β L At that time, with β G As β increases from 0.1 to 1.0, the size of the failed nodes affected by LLRM in the general generative network model shows an increasing trend; when β L Then, with β G As the β value increases from 0.1 to 1.0, the number of failed nodes affected by LLRM in the general generative network model decreases. Different β values... L At a certain value, the effect of LLRM varies with β. G The changes show completely opposite trends, indicating that the effects of GLRM and LLRM are both mutually reinforcing and mutually inhibiting. (Compare β...) G The data, i.e., the results of LLRM acting independently, also lead to the above conclusion. The reason for this is due to β... LWhen the load is relatively small, the redundancy capacity of nodes in the general generative network model to resist local load surges is... When β is relatively small, the local load redistribution caused by a faulty node can easily lead to overload and failure of its neighboring nodes, allowing the fault to spread gradually from the fault point to the fully general generative network model under topological constraints. G When the β value is small, the cascading disruption of the general generative network model by GLRM propagates more rapidly than that of LLRM due to the lack of topological constraints; as β increases... G As β increases, the destructive effect of GLRM on general generative network models gradually decreases, allowing LLRM to gradually exhibit its destructive effects. Therefore, β L When the value is small, the violation of the general generative network model by LLRM increases with β. G It increases with the increase of β. L As β increases, the effect of LLRM decreases. G The changes exhibit a trend opposite to the above phenomena. Further analysis reveals that this is because the impact of LLRM primarily depends on cascading faults caused by localized energy surges, β G The larger the value, the more likely it is that multiple adjacent nodes need to fail simultaneously to create a combined energy surge, causing cascading failures in its neighboring nodes. However, the impact range of GLRM is relatively dispersed, leading to more failures in non-adjacent regions. G The larger the β value, the fewer faulty nodes the GLRM causes, and the less likely it is to form node faults concentrated in a localized area. On the one hand, if β... L When β is relatively small, these dispersed fault nodes can cause multiple regions to trigger local cascading simultaneously, amplifying the impact of LLRM and demonstrating the promoting effect of GLRM on LLRM; on the other hand, if β L Larger, dispersed fault nodes are unlikely to generate the localized concentrated energy surge required for LLRM. Instead, these dispersed fault nodes will consume local energy in stages, reducing the energy surge efficiency under LLRM, exhibiting a variation with β. G The increase in the size of the failed nodes under LLRM results in a reduction in the size of the failed nodes, which is the effect of GLRM in suppressing the effect of LLRM.

[0086] See Figure 2Part (b) shows that the effects of LLRM on GLRM are both mutually reinforcing and mutually inhibiting. When LLRM causes more node failures in the general generative network model in the initial stage, its promoting effect on GLRM is manifested as follows: the more nodes that fail in the general generative network model at the current stage, the greater the change in the overall topology of the general generative network model, resulting in more chaotic energy flow, which is more conducive to GLRM causing more network damage. Its inhibiting effect on GLRM is manifested as follows: as the number of failed nodes increases, the total energy flow in the general generative network model decreases, which is equivalent to increasing the redundancy capacity of the remaining nodes, weakening the ability of GLRM to cause subsequent network damage. Because LLRM is constrained by the network topology, its effects need to be gradually revealed; therefore, the overall promoting effect of LLRM on GLRM is greater than its inhibiting effect.

[0087] according to Figure 2 Part (c) shows that the multi-dynamic coupling cascade effect under TTIP makes the network more vulnerable. Combined with... Figure 3 A comparative analysis of the cascading failure node sizes in TTIP reveals that the cascading effects under TTIP generally show a higher degree of consistency with the trend of cascading size variation with parameters under GLRM influence. It can be argued that the main effect of cascading failure in TTIP is caused by GLRM, while LLRM primarily manifests its effect by promoting the influence of GLRM, only when β... L The effects of LLRM become more pronounced when the cascade size is small. The overall effect of the two dynamic mechanisms in the cascade process is that they compete with each other when they each have a strong cascade diffusion capacity, and cooperate with each other when their cascade diffusion capacity is relatively weak, jointly causing greater cascade damage.

[0088] in, Figure 3 In parts (a1) to (a3) ​​and parts (b1) to (b3), “LLRM” refers to the time-overloaded node affected by LLRM, “GLRM” refers to the time-overloaded node affected by GLRM, and “Total” refers to the total number of failed nodes in the general generative network model.

[0089] Correspondingly, refer to Figure 4 This shows the TTIP network within the xx region corresponding to different reallocation parameter combinations β. G and β L The simulation results are as follows; based on Figure 4 Compare parts (a) to (c). Figure 2 From (a) to (c), the simulation results of the general generative network model under the corresponding conditions show that the characteristics and trends of the simulation data of the route network in the xx region meet the above analysis conclusions, proving the effectiveness of the proposed model for real network analysis.

[0090] Part Two: Cascade Failure Characteristics Analysis of TTWT like Figure 5 As shown in parts (a) and (b), TTWT can be understood as setting LLRM and GLRM to have the same capacity parameter, β, in the TTIP model, compared to TTIP. G =β L Furthermore, the redundant capacity of nodes under LLRM and GLRM can be shared, i.e. and Both can bear It can also carry .

[0091] In scale-free networks, nodes with large local and global loads have a high overlap rate. Therefore, the distribution of heavily loaded nodes is highly consistent in the models corresponding to TTWT, LLRM, and GLRM. Both LLRM and GLRM follow the rule that nodes with larger loads will receive more load or have a higher probability of redistribution. TTWT also conforms to this rule. These factors cause the cascading effects of multiple dynamic mechanisms in TTWT to highly overlap with the effects of individual LLRM or GLRM mechanisms, preventing the coupling effect of LLRM and GLRM from being realized. Ultimately, the cascading effect is no better than that of a single dynamic mechanism.

[0092] Compared to TTIP, TTWT enhances the network's resilience to cascading failures caused by multiple dynamic mechanisms. The destructive impact of TTWT on the network is similar to that of LLRM, lower than that of GLRM alone, and even lower than that of TTIP under the same parameters. This is because cascading failures under LLRM are mainly affected by the load redistribution caused by the simultaneous failure of relatively concentrated adjacent nodes in the same phase. When the size is small, the effect of LLRM can cause localized cascading diffusion without the help of GLRM; while when At higher loads, GLRM cannot help LLRM create a locally concentrated load redistribution impact. Therefore, under TTWT, the cascading effect of LLRM is less affected by GLRM. However, the total network load under GLRM is proportional to the existing network size. When LLRM acts on the network, the faulty nodes caused by LLRM will reduce the total network load, which is equivalent to indirectly increasing the redundancy capacity of the remaining nodes in the network, thereby reducing the impact of GLRM on network cascading failures.

[0093] Meanwhile, the redundant capacity of the two loads can be shared in TTWT, while the capacities of the two loads are independent in TTIP. Combined with LLRM... The cascading effect is insufficient at larger scales. The redundancy capacity of the model corresponding to LLRM provides additional capacity to withstand the energy redistribution under overload in the model corresponding to GLRM. This can be equivalently considered as increasing the redundancy capacity of nodes to resist the redistribution energy under GLRM. For example, in TTIP, the redistribution load under the model corresponding to LLRM can be supported by its redundancy capacity, but the redundancy capacity under the model corresponding to GLRM is insufficient to support the redistribution load under GLRM, thus causing node failure. However, in the same situation in TTWT, the remaining redundancy capacity can still be used to help withstand the excessive redistribution load under GLRM, ultimately preventing node cascading failure.

[0094] Both of these effects reduce the cascading effect under GLRM, ultimately causing the cascading failure of TTWT to be closer to the effect of LLRM acting alone, and lower than the cascading effect of TTIP under the same capacity parameters. Clearly, from a network protection perspective, when these multiple dynamic mechanisms coexist in the network, designing node capacity to simultaneously accommodate both types of loads can more effectively resist the cascading effects of the coupling mode of multiple dynamic mechanisms in TTIP.

[0095] Part Three: Cascade Failure Characteristics Analysis of OTCE See Figure 6 From parts (a) and (b), we can conclude that: as β GWith the addition of [unspecified element], the simulation results under OTCE and GLRM generally show a consistent trend, which can be explained by the similarity of their models. The difference lies in that OTCE couples the influence of LLRM on top of GLRM. Simulation results show that OTCE has greater cascade vulnerability than GLRM, indicating that the addition of LLRM enhances the cascade diffusion effect of GLRM. Unlike TTIP, in OTCE, the influence of LLRM on GLRM is more of a promoting effect, especially when the capacity parameter is set appropriately, the promoting effect of LLRM on GLRM shows a non-linear increase. Compared to the load superposition result after the diffusion of two loads according to their own dynamic mechanisms in TTWT, the coupling effect of a single load under multiple dynamic mechanisms in OTCE obviously has a stronger cascade destructive effect. This is because, in both TTIP and TTWT, nodes bear two loads and correspondingly have the capacity to bear both loads. However, nodes in OTCE only have the capacity for one load setting. Therefore, adjacent nodes of a failed node are not only affected by global load redistribution adjustments, but also by local load redistribution impacts from the failed node itself. This means they are simultaneously affected by two redistributed loads, leading to a higher risk of cascading failures for adjacent nodes. Furthermore, the scale of this cascading failure is proportional to the scale of the failed node in the previous stage. Therefore, overall, OTCE represents the most cascading threat multi-dynamic mechanism coupling model for network security.

[0096] Based on the above description, different dynamic mechanisms have different cascading processes and effects, and the combined effect of multiple dynamic mechanisms can produce more complex coupling effects. This application models the topology of infrastructure systems through complex networks and studies the cascading effects under three multi-dynamic mechanism coupling modes based on the dynamic mechanisms in two classic load-capacity models: LLRM and GLRM. Simulation results show that in TTIP, the two mechanisms both promote and inhibit each other, but overall enhance network vulnerability. In TTWT, the sharing of redundant capacity improves the network's resistance to cascading effects. In OTCE, the coupling of LLRM shows a promoting effect on GLRM, exacerbating the impact of cascading failures. Therefore, the dynamic mechanisms involved in TTWT can be selected here to manage the route network within the xx region.

[0097] Thus, this application reveals for the first time the fundamental transformation of the interaction between the two dynamic mechanisms, LLRM and GLRM, under the three coupling modes of TTIP, TTWT, and OTCE. Based on this discovery, this application, by selectively applying the three coupling modes of TTIP, TTWT, and OTCE, achieves for the first time the proactive bidirectional control of the cascading effects caused by multi-mechanism coupling in complex networks, that is, it can both reinforce and compress, providing a unified and efficient technical framework for robustness assessment, survivability design, and disaster scenario simulation of complex networks.

[0098] Correspondingly, the coupling effects of multiple dynamic mechanisms further complicate the dynamic behavior of complex giant systems, making it difficult to effectively predict and extrapolate the effects of single dynamic mechanisms. Moreover, different coupling modes of multiple dynamic mechanisms have varying effects on network cascading, potentially enhancing network resilience or making it more vulnerable, and may exhibit mutually reinforcing or inhibiting transformations under different parameters. Therefore, future research should focus on the coupling of two dynamic mechanisms within the load-capacity model. Further in-depth research and exploration are needed to investigate the coupling effects of more types and categories of dynamic mechanisms.

[0099] 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.

[0100] 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.

[0101] In the several embodiments provided in this application, it should be understood that the disclosed methods can be implemented in other ways.

[0102] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0103] The features disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0104] The above description is merely an embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method of managing a network-based infrastructure system based on multi-dynamics coupling, characterized in that, The method includes: Model the infrastructure system in the preset area that is disturbed in the current time period to obtain the network to be processed; Under the combined action of the local load redistribution mechanism LLRM and the global load redistribution mechanism GLRM, a coupled model of multiple dynamic mechanisms corresponding to the network to be processed is constructed. Dynamic simulations were performed on each of the multiple dynamic mechanism coupling models to obtain the simulation results of cascade failure characteristics under each multiple dynamic mechanism coupling model. Based on the assessment objectives of the infrastructure system in the current period, and based on the simulation results of cascading failure characteristics, the target coupling model of the infrastructure system is determined from multiple multi-dynamic mechanism coupling models, and the infrastructure system is managed based on the dynamic mechanisms within the target coupling model.

2. The method of claim 1, wherein, Modeling the infrastructure system with disturbances within a preset area during the current time period yields the network to be processed that has experienced cascading failures, including: The units and relationships operating within the infrastructure system are abstracted as network nodes and connection edges, respectively. The state of network nodes and connection edges is labeled and the relationships are quantified to obtain the network to be processed.

3. The method according to claim 1 or 2, characterized in that, Multiple coupled models of various dynamic mechanisms are included: the Two-Type Load-Two-Dynamic Mechanism Independent Propagation Model (TTIP); under the combined action of the Local Load Redistribution Mechanism (LLRM) and the Global Load Redistribution Mechanism (GLRM), a multiple coupled model of various dynamic mechanisms corresponding to the network to be processed is constructed, including: Obtain faulty nodes in the network to be processed Transferred to node under LLRM action Local redistribution of load and based on local load redistribution For nodes Adjust the current load to obtain the node New local load ; Based on the fault nodes of the network to be processed Global load balancing performed under GLRM results in node... New global load ; exist or Representation Nodes In the event of overload failure, a TTIP is constructed to cascade the fault propagation based on the LLRM for local load and the GLRM for global load. in, and They are nodes The total initial capacity under LLRM and GLRM.

4. The method according to claim 1 or 2, characterized in that, Multiple coupled models of various dynamic mechanisms are included: a two-load-two-dynamic-mechanism co-operation model (TTWT); under the combined action of the local load redistribution mechanism (LLRM) and the global load redistribution mechanism (GLRM), multiple coupled models of various dynamic mechanisms corresponding to the network to be processed are constructed, including: Obtain faulty nodes in the network to be processed Transferred to node under LLRM action Local redistribution of load and based on local load redistribution For nodes Adjust the current load to obtain the node New local load ; Based on the fault nodes of the network to be processed Global load balancing performed under GLRM results in node... New global load ; exist Representation Nodes In the event of overload failure, the corresponding TTWT is constructed for cascading fault propagation based on LLRM for local load and GLRM for global load. in, For nodes The initial total capacity set under the combined action of LLRM and GLRM, and , and All of these are adjustable capacity parameters.

5. The method according to claim 1 or 2, characterized in that, Multiple multi-dynamic mechanism coupling models are proposed, including: a load-two-dynamic-mechanism coupling effect model (OTCE); and a multi-dynamic mechanism coupling model corresponding to the network to be processed, constructed under the combined action of the local load redistribution mechanism (LLRM) and the global load redistribution mechanism (GLRM), including: Integrate faulty nodes in the network to be processed Transferred to node under LLRM action Local redistribution of load And nodes after global load balancing based on GLRM for other non-failed nodes in the network to be processed. Background load: , obtain node New current load: ; in, The number of nodes in the network to be processed; For nodes As a receiving unit, it receives data from other non-failed [receivers / receivers]. The load stream received by each node; For all node pairs in the network to be processed The load passes through the nodes The sum of the loads; exist Representation Nodes In the event of overload failure, the corresponding OTCE is constructed by cascading fault propagation based on LLRM for local load and GLRM for global load. in, For nodes The initial total capacity set under the combined coupling effect of LLRM and GLRM.

6. The method according to claim 1, characterized in that, An infrastructure system is identified as having disturbances if an internally operating unit is detached from the main body of the infrastructure system, and / or if an internally operating unit experiences an overload failure.

7. The method according to claim 1, characterized in that, Dynamic simulations were performed on each of the multiple multi-dynamic mechanism coupling models to obtain simulation results of cascading failure characteristics under each multi-dynamic mechanism coupling model, including: For each multi-dynamic mechanism coupling model, the initial topology and initial load distribution of the multi-dynamic mechanism coupling model are mapped for coupling relationship and the failure threshold is calibrated to obtain the simulation initialization configuration file of the multi-dynamic mechanism coupling model; For the simulation initialization configuration file, cascading propagation iterations are performed within discrete time steps, and the load overload node and failure propagation range are recorded at each time step to obtain the cascading failure characteristic simulation results of each of the multi-dynamic mechanism coupled models, including the failure timing and cascading scale.

Images