Emergency response tissue toughness evaluation method and system based on complex network

Abstract

The invention belongs to the technical field of emergency response, and particularly discloses an emergency response tissue toughness evaluation method and system based on a complex network. Comprising the following steps: constructing an organizational relation matrix according to actual response data and emergency plan data so as to establish a relation and an interaction mechanism among subjects in an emergency response process; constructing an organization network of four stages of pre-disaster plan, emergency rescue, recovery and reconstruction and post-disaster plan according to the organization relation matrix; the node importance is evaluated; the method comprises the following steps: simulating a network operation condition under a single attack strategy to perform toughness evaluation on a network organization of residual efficiency characterization emergency response; coupling a plurality of network attack strategies by utilizing an entropy weight method, further simulating different high-hazard risk scenes, and calculating the weight of each attack strategy by calculating information entropy; and evaluating the network efficiency under various attack scenes to obtain the overall toughness level of the emergency response organization in each stage. The method can clearly reflect the emergency response capability and recovery capability of the tissue in different stages.

Description

Technical Field

[0001] The present invention belongs to the field of emergency response technology, and more specifically, relates to a complex network-based emergency response organization resilience evaluation method and system. Background Art

[0002] With the rapid development of society and the advancement of globalization, various emergencies and disasters are occurring more frequently. How to quickly and effectively respond to these crises and protect people's lives and property has become a critical issue for governments and businesses at all levels. In emergency response, organizational resilience—the ability to recover and adapt after encountering an emergency—is directly related to the efficiency and effectiveness of the response. Therefore, how to scientifically and objectively assess an organization's emergency response resilience has become a key research and practice topic.

[0003] Existing methods for assessing the resilience of emergency response organizations mostly focus on qualitative assessments, relying on subjective judgments based on expert experience. This makes it difficult to comprehensively and accurately reflect an organization's emergency response capabilities. Traditional assessment methods may overlook the structural influence of organizational networks and fail to fully consider the relationships between various nodes within the organization and their role in emergency response. This results in significant limitations in practical application of existing methods, making it difficult to provide a sufficient reference for decision makers.

[0004] With the rise of complex network theory, a growing number of studies have begun applying network analysis methods to assess organizational resilience. Complex network theory offers new insights into understanding and modeling the connections between departments and individuals within an organization. By viewing an organization as a complex network, researchers can quantitatively assess its emergency response capabilities in the face of emergencies from multiple perspectives, such as network structure and node importance. However, existing complex network-based organizational resilience assessment methods often lack simulations for diverse attack strategies and node failure scenarios, making them incapable of comprehensively assessing an organization's response and recovery capabilities in the face of diverse emergencies or attacks. Furthermore, current research on emergency response organizations faces the following shortcomings: First, in-depth research on differences in organizational network structure across different response phases is insufficient; second, static network structure analysis methods are insufficient to describe the dynamic complexity of organizational resilience or the risk context within which organizational resilience exists. Complex network methods have not yet been integrated with organizational resilience assessment systems from a dynamic perspective, resulting in a lack of network-based organizational resilience assessment methods; third, clear comparisons and assessments of organizational resilience across different response phases are lacking. Existing research on emergency response organizational networks primarily relies on static network structure analysis, offering structural explanations for the potential black box nature of emergency response networks.

[0005] Therefore, there is an urgent need for a complex network-based emergency response organization resilience evaluation method that can comprehensively and accurately evaluate the organization's emergency response capabilities and resilience under different attack scenarios through scientific models and analysis tools, and provide effective data support for emergency management decisions. Summary of the Invention

[0006] In response to the above-mentioned defects or improvement needs of the prior art, the present invention provides an emergency response organization resilience evaluation method and system based on complex networks. It proposes a new organizational resilience evaluation system by deeply analyzing the structural differences of organizational networks in different response stages, combining dynamic network models and complex network theory. Specifically, the present invention comprehensively evaluates the resilience performance of the organization in the entire emergency response process, which is divided into four stages: pre-disaster plan, emergency rescue, recovery and reconstruction, and post-disaster plan, from multiple perspectives such as network effectiveness, node importance, and attack strategy. It can clearly reflect the emergency response capability and recovery capability of the organization at different stages. This method can be widely used in emergency management, disaster response, enterprise risk assessment, and various organizational management fields that need to improve emergency response capabilities.

[0007] To achieve the above objectives, according to one aspect of the present invention, a complex network-based emergency response organization resilience evaluation method is proposed, comprising the following steps:

[0008] Step 1: Based on the actual response data and emergency plan data, an organizational relationship matrix is ​​constructed to establish the relationship and interaction mechanism between various subjects in the emergency response process;

[0009] Step 2: Construct an organizational network for the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix;

[0010] Step 3: Use the weighted degree, betweenness centrality, closeness centrality, and K-core node importance metrics to evaluate node importance;

[0011] Step 4: By simulating the network operation under a single attack strategy, the resilience of the network organization representing the emergency response is evaluated;

[0012] Step 5: Use the entropy weight method to couple multiple network attack strategies, simulate different high-risk scenarios, and calculate the weight of each attack strategy by calculating the information entropy;

[0013] Step six: Evaluate the network effectiveness under various attack scenarios to determine the overall resilience level of the emergency response organization at each stage.

[0014] As further preferred, step one includes the following steps:

[0015] (11) According to different stages of emergency response, a multi-source mixed data collection method is used to collect actual response data and emergency plan data;

[0016] (12) Integrate, clean and pre-process the text data of the actual response data and emergency plan data obtained;

[0017] (13) The organizational network is constructed through the relationship extraction method based on text analysis, and the information extraction and relationship expression between organizations are realized;

[0018] (14) The co-occurrence frequency of organizations is used as the edge weight of organizational relationships to obtain the organizational relationship matrix, and the relationship and interaction mechanism between various subjects in the emergency response process are established.

[0019] As further preferred, step 2 includes the following steps:

[0020] (21) Based on the organizational relationship matrix, the organizational network of the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning was constructed, and the network was visualized using Gephi software;

[0021] (22) Community division and visualization are performed through Modularity Class modularity clustering. Node affiliation is divided by node color, and node size is described according to weighted degree.

[0022] (23) Through the topological structure of the organizational network, visualization of its components and measurement analysis of its overall structural indicators, we can find the structural characteristics and stage differences of the organizational response network.

[0023] As further preferred, step three includes the following steps:

[0024] (31) The importance of nodes is evaluated using weighted degrees. The greater the weighted degree, the greater the influence of the node and the more important the node.

[0025] (32) Betweenness centrality is used to evaluate the importance of nodes. The higher the betweenness centrality, the more important the node is.

[0026] (33) Closeness centrality is used to evaluate the importance of nodes. The greater the closeness centrality of a node, the more central the node is in the network and the closer it is to other nodes in the network.

[0027] Preferably, the calculation model of the weighted degree includes:

[0028]

[0029] Where k i is the weighted degree of node i, h ij is the edge weight between node i and its neighbor node j;

[0030] Preferably, the calculation model of betweenness centrality includes:

[0031]

[0032] Where BC(i) is the betweenness centrality of node i, V is the set of nodes in the network, and n st is the number of shortest paths between nodes s and t, n st (i) is the number of shortest paths between points s and t that pass through node i;

[0033] Preferably, the calculation model of the closeness centrality includes:

[0034]

[0035] Where CC(i) is the closeness centrality of node i, N is the total number of nodes, and d ij is the shortest distance between nodes i and j.

[0036] As further preferred, step 4 includes the following steps:

[0037] (41) By systematically evaluating the importance of nodes, we can simulate network behavior under different attack strategies;

[0038] (42) By simulating the network operation under a single attack strategy, the network effectiveness reflected by the remaining organizational network is evaluated to represent the organizational resilience of the emergency response network;

[0039] (43) According to the attack strategy, the network performance index value is reduced to 0, that is, the order of network complete collapse is analyzed, and the influence of different attack strategies on network performance is sorted in descending order.

[0040] As a further preferred embodiment, the simulation of network behaviors under different attack strategies includes:

[0041] Simulate scenarios where nodes are attacked, deleted, or destroyed, removing the attacked nodes and their connections, leading to fragmentation and performance loss of the organization's network.

[0042] As a further preferred embodiment, in step 5, the calculation formula of the entropy weight method includes:

[0043]

[0044] Where, X ij is the standardized network performance index data, T ij is the normalized index value, e i is the information entropy calculation value, w iis the weight of each attack strategy obtained by the entropy weight method.

[0045] As a further preferred embodiment, step five further includes:

[0046] (51) The entropy weight method is used to couple multiple network attack strategies and simulate different high-risk scenarios. After standardizing the data, the weight of each attack strategy is obtained by calculating the information entropy.

[0047] (52) Since the robustness, transmission and recovery of organizational networks are positively correlated with organizational resilience, and the clustering is negatively correlated with organizational resilience, based on the network effectiveness evaluation results and their correlation with organizational resilience, the organizational resilience evaluation is achieved by simply averaging the network effectiveness evaluation results.

[0048] As a further preferred embodiment, in step 6, the emergency response organization resilience evaluation index based on network effectiveness evaluation is as follows:

[0049]

[0050] Where, F j represents the network performance evaluation value, w i is the attack strategy weight, f i is the performance index value; R represents the tissue toughness evaluation value, h i Use + / - according to the positive or negative impact of performance on toughness.

[0051] Based on any of the above embodiments or a combination of multiple embodiments, according to another aspect of the present invention, a complex network-based emergency response organization resilience evaluation system is provided, comprising:

[0052] The first main control module is used to build an organizational relationship matrix based on actual response data and emergency plan data to establish the relationship and interaction mechanism between various subjects in the emergency response process;

[0053] The second main control module is used to build an organizational network for the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix;

[0054] The third main control module is used to evaluate the importance of nodes using the node importance metrics of weighted degree, betweenness centrality, closeness centrality, and K-core;

[0055] The fourth main control module is used to evaluate the resilience of the network organization that represents the emergency response by simulating the network operation under a single attack strategy;

[0056] The fifth main control module is used to couple multiple network attack strategies using the entropy weight method, and then simulate different high-risk scenarios to obtain the weight of each attack strategy by calculating the information entropy;

[0057] The sixth main control module is used to evaluate the network effectiveness under various attack scenarios and obtain the overall resilience level of the emergency response organization at each stage.

[0058] In general, the above technical solutions conceived by the present invention have the following technical advantages compared with the existing technology:

[0059] 1. This paper proposes a new organizational resilience evaluation system by deeply analyzing the structural differences of organizational networks at different response stages and combining dynamic network models with complex network theory. Specifically,

[0060] The present invention uses the relationship extraction method to realize the construction of the organizational network, takes the co-occurrence frequency of the organization as the edge weight of the organizational relationship, obtains the organizational relationship matrix, and establishes the relationship and interaction mechanism between the various subjects in the emergency response process;

[0061] This paper takes into account multiple perspectives, such as network effectiveness, node importance, and attack strategies. Based on the full lifecycle perspective of emergency response, and combining the dual functional attributes of "pre-disaster planning" and "post-disaster optimization" of emergency plans, it integrates the "pre-disaster-during-post-disaster" full cycle with the "plan-practice-optimization" closed-loop link. Through text mining and complex network methods, it constructs a four-stage emergency response organization collaborative network of "pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning", thereby mining deeper implicit information of organizational networks.

[0062] The present invention uses the node importance measurement of weighted degree, betweenness centrality, closeness centrality and K-core to evaluate the importance of nodes, which is a new method for evaluating the importance of actual emergency response organizations.

[0063] The present invention simulates network operation under various single attack strategies (such as random attacks, intentional attacks based on node weighted degree, betweenness centrality, closeness centrality, K-core, etc.) with high fidelity. By precisely setting attack parameters and network environment variables, the simulation results can truly reflect the actual performance of the emergency response network under different attack scenarios.

[0064] This paper uses an entropy weighting method to couple multiple attack strategies and scientifically calculate weights, fully accounting for the diverse scenarios encountered in actual emergency responses. The simulation captures the sequence of different attack types, changes in attack intensity, and interactions between them. This provides strong support for relevant departments in developing scientifically sound emergency response strategies, helping to improve the efficiency and effectiveness of emergency responses and reduce potential losses.

[0065] The present invention evaluates the network performance under various attack scenarios, mainly including the dynamic evaluation of the organizational resilience of the organizational network in terms of robustness, transmission, recovery, and aggregation, and obtains the overall resilience level of the emergency response organization at each stage. It can provide support for the construction of organizational nodes and the optimization of organizational paths, and has broad application expansion potential. It is not only applicable to various types of emergency response organization networks, such as disaster relief networks, public health emergency networks, etc., but can also be extended to other complex network systems, such as transportation networks, communication networks, etc. By dynamically evaluating the organizational resilience of networks in different fields under attack scenarios, it can provide scientific and effective methods for network planning, construction, and management in various industries, helping to improve the ability and level of the entire society to respond to emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 Schematic diagram of a framework of a complex network-based emergency response organization resilience evaluation method according to an embodiment of the present invention;

[0067] Figure 2 This is a flowchart of an emergency network construction process according to an embodiment of the present invention;

[0068] Figure 3 This is the organizational network topology structure of the pre-disaster emergency plan, emergency rescue, recovery and reconstruction, and post-disaster emergency plan involved in the embodiment of the present invention;

[0069] Figure 4 This is a diagram showing the results of a network attack simulation organized by a pre-disaster emergency plan according to an embodiment of the present invention;

[0070] Figure 5 This is a diagram showing the results of a network attack simulation organized by a post-disaster emergency plan according to an embodiment of the present invention;

[0071] Figure 6 This is a diagram showing the simulation results of a network attack on an emergency rescue organization involved in an embodiment of the present invention;

[0072] Figure 7 This is a diagram showing the results of a network attack simulation for restoring and reconstructing an organization involved in an embodiment of the present invention;

[0073] Figure 8 is a schematic diagram of the organizational network effectiveness evaluation results involved in an embodiment of the present invention;

[0074] Figure 9 Schematic diagram of the tissue toughness evaluation results according to an embodiment of the present invention. DETAILED DESCRIPTION

[0075] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0076] like Figure 1 and Figure 2 As shown, an embodiment of the present invention provides an emergency response organization resilience evaluation method based on a complex network, and the specific steps are as follows: S1: for different stages of emergency response, a multi-source mixed data collection method is adopted, and on the basis of the data collection of emergency plans and actual responses, the acquired text data is pre-processed by integration, cleaning, etc.; S2: text analysis technology is used to construct a noun organization dictionary to realize information extraction and expression of association relationships between organizations, and the relationship extraction method is used to realize organizational network construction, and the co-occurrence frequency of organizations is used as the edge weight of organizational relationships to obtain an organizational relationship matrix, and establish the relationship and interaction mechanism between various subjects in the emergency response process; S3: according to the organizational relationship matrix, an organizational network of four stages, namely, pre-disaster plan, emergency rescue, recovery and reconstruction, and post-disaster plan, is constructed, and then more Deep-level implicit information of organizational networks; S4: Use weighted degree, betweenness centrality, closeness centrality, and K-core node importance metrics to evaluate node importance, and then realize organizational resilience evaluation; S5: By simulating the network operation under a single attack strategy, at this time, the network effectiveness reflected by the remaining organizational network is characterized to evaluate the organizational resilience of the emergency response network; S6: Use the entropy weight method to couple multiple network attack strategies, and then simulate different high-risk scenarios. After standardizing the data, the weight of each attack strategy is obtained by calculating the information entropy; S7: Evaluate the network effectiveness under multiple attack scenarios, mainly including the dynamic evaluation of the organizational resilience of the robustness, transmission, recovery, and aggregation of the organizational network, and obtain the overall resilience level of the emergency response organization at each stage. Specifically, it includes the following steps:

[0077] Step 1: Using a heavy rainstorm disaster in a certain area as the background, a multi-source hybrid data collection method was used for different stages of emergency response. Based on the data collected from the emergency plan and actual response, the acquired text data was pre-processed by integration, cleaning, etc.

[0078] Step 2: Use text analysis technology to extract information and express associations between organizations. Use the relationship extraction method based on text analysis to construct an organizational network. Use the co-occurrence frequency of organizations as the edge weight of organizational relationships to obtain an organizational relationship matrix and establish the relationships and interaction mechanisms between various entities in the emergency response process.

[0079] Step 3: Based on the organizational relationship matrix, construct the organizational network of the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning, and then explore the deeper implicit information of the organizational network;

[0080] Step 4: Use the node importance metrics of weighted degree, betweenness centrality, closeness centrality, and K-core to evaluate the node importance, and then achieve organizational resilience evaluation;

[0081] Step 5: By simulating the network operation under a single attack strategy, the network effectiveness reflected by the remaining organizational network is evaluated to represent the organizational resilience of the emergency response network;

[0082] Step 6: Use the entropy weight method to couple multiple network attack strategies. By simulating different high-risk scenarios, after standardizing the data, the weight of each attack strategy is calculated by calculating the information entropy to evaluate the network effectiveness under the attack scenario, mainly including the dynamic evaluation of the organizational resilience of the organizational network in terms of robustness, transmission, recovery, and aggregation, and the overall resilience level of the emergency response organization at each stage is obtained.

[0083] Furthermore, in the present invention, the described step 1 specifically includes:

[0084] Step S11: A multi-source hybrid data collection method is used for different stages of the emergency response. Data sources are mainly divided into two categories: one is the actual organizational response data, which comes from the official government website, including special news reports, emergency press conferences, and post-disaster reconstruction press conferences; the other is emergency plan data, which comes from official documents for pre-disaster and post-disaster emergency plans. Taking a case study of a heavy rainstorm in a certain area, the data sources are mainly divided into two categories: one is the actual organizational response data, which comes from the official government website, and the other is the emergency plan data.

[0085] Step S12: The entire actual response process is divided into two phases: emergency rescue and recovery and reconstruction. Using the time of disaster occurrence as the dividing point, the emergency plan is divided into two phases: pre-disaster plan and post-disaster plan. This divides the entire emergency response process into four phases: pre-disaster plan, emergency rescue, recovery and reconstruction, and post-disaster plan. Based on differences in data sources, the emergency response process can be divided into two categories: planned response based on the emergency plan and actual response. This allows for the exploration of the characteristic differences between planned and actual responses. A total of 401 texts were collected for the actual response process, including 226 from the emergency rescue phase and 175 from the recovery and reconstruction phase.

[0086] Step S13: Based on key time nodes, the emergency response process can be divided into four stages: "pre-disaster plan - emergency rescue - recovery and reconstruction - post-disaster plan". That is, the entire emergency response process is divided into four stages: pre-disaster plan, emergency rescue, recovery and reconstruction, and post-disaster plan.

[0087] Step S14: Based on the data collection of emergency plans and actual responses, the acquired text data is pre-processed by integration, cleaning, etc. The text pre-processing operation specifically filters out stop words and special characters, and retains nouns as potential keywords.

[0088] In step 2, text analysis technology is used to construct a noun organization dictionary to extract information and express association relationships between organizations. The relationship extraction method is used to construct the organizational network. The co-occurrence frequency of organizations is used as the edge weight of the organizational relationship to obtain the organizational relationship matrix. The relationship and interaction mechanism between the various subjects in the emergency response process are established, such as Figure 3 As shown, specifically including:

[0089] Step S21: Further improve and supplement the organization name dictionary, build the organization network through the relationship extraction method based on text analysis, and realize the information extraction and relationship expression between organizations

[0090] Step S22: By constructing a noun organization dictionary, information extraction and association expression between organizations are realized. The organizational network is constructed using the relationship extraction method. The co-occurrence frequency of organizations is used as the edge weight of the organizational relationship to obtain the organizational relationship matrix, and the relationship and interaction mechanism between the various subjects in the emergency response process are established;

[0091] In step 3, the organizational network of the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning is constructed based on the organizational relationship matrix, and then the deeper implicit information of the organizational network is mined, including:

[0092] Step S31: constructing an organizational network for the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix, and visualizing the network using Gephi software;

[0093] Step S32: Visualize the community division by modularity class clustering. Node affiliation is divided by node color, and node size is described according to weighted degree. That is, the larger the node in the graph, the higher the weighted degree.

[0094] Step S33: Explore the structural characteristics and stage differences of the organizational response network through the topological structure of the organizational network, visualization of components and measurement analysis of overall structural indicators.

[0095] In step 4, the node importance metrics of weighted degree, betweenness centrality, closeness centrality, and K-core are used to evaluate the node importance, thereby achieving organizational resilience evaluation, specifically including:

[0096] S41: Organizational resilience evaluation is based on the evaluation of organizational network effectiveness under attack strategies. Different intentional attack strategies are based on different node importance metrics, so node importance must be evaluated before network attack simulation can be realized. The network effectiveness under attack scenarios must be evaluated before organizational resilience can be realized.

[0097] S42: Node importance measurement. The organizational resilience evaluation method in the present invention is based on the evaluation of organizational network effectiveness under attack strategies. Different intentional attack strategies are adopted in the present invention, and the selection of these strategies is based on the importance measurement of each node. In order to achieve this, it is necessary to quantitatively evaluate the criticality of each node. Commonly used node importance measurement indicators include node degree, betweenness centrality, closeness centrality, etc. These metrics reflect the importance and role of each node in the network. The specific node importance measurement indicators are as follows:

[0098] S43: Weighted degree. In weighted networks, weighted degree is the most intuitive and commonly used metric for measuring node importance. A node's weighted degree is the sum of the weights of all edges connected to that node, calculated as shown in Equation (1). The greater a node's weighted degree, the greater its influence and the more important it is.

[0099]

[0100] Where k i is the weighted degree of node i, h ij is the edge weight between node i and its neighbor node j.

[0101] S44: Betweenness centrality. Betweenness centrality describes the connection and transit capabilities of a node, and is calculated as Equation (2). Betweenness centrality evaluates the importance of a node from a global network perspective. Nodes with high betweenness centrality play important roles as "intermediaries," "bridges," and "hubs" in the network.

[0102]

[0103] Where BC(i) is the betweenness centrality of node i, V is the set of nodes in the network, and n st is the number of shortest paths between nodes s and t, n st (i) is the number of shortest paths between points s and t that pass through node i.

[0104] S45: Closeness centrality. Closeness centrality reflects the accessibility and closeness of the node to other nodes. The calculation formula is formula (3). The greater the closeness centrality of a node, the more central the node is in the network and the closer it is to other nodes in the network.

[0105]

[0106] Where CC(i) is the closeness centrality of node i, N is the total number of nodes, and d ij is the shortest distance between nodes i and j.

[0107] S46: K-core. The K-core algorithm considers the overall structure of the network, not only the node's degree but also its location within the network. In a network, a node located in the inner core typically has a high level of influence despite having a low degree. On the other hand, a node located at the edge of the network, despite having a high degree, has relatively limited influence. The K-core decomposition algorithm, based on the node's degree, gradually removes nodes with low degrees, meaning those located at the edge of the network, achieving a hierarchical division of network nodes based on their degree.

[0108] S47: Key organizations usually refer to organizational nodes with high influence, strong resource allocation capabilities, or core coordination functions. These organizations occupy an important position in the topological structure of the network. Their failure or functional impairment may have a significant impact on the stability and efficiency of the entire emergency response system. Different node importance evaluation indicators have different focuses. In order to more comprehensively identify the "roles" and functional positioning of key organizations in the network from multiple perspectives, this paper uses weighted degree, betweenness centrality, proximity centrality, and K-core indicators, covering node importance measurement methods based on neighbor nodes, node locations, and paths, to identify and analyze key organizations in the emergency response organization network.

[0109] In step 5, by simulating the network operation under a single attack strategy, the network effectiveness of the remaining organizational network is evaluated to represent the organizational resilience of the emergency response network, including:

[0110] S51: Attack Simulation. By systematically assessing node importance, we can simulate network behavior under different attack strategies. Specifically, we simulate scenarios where nodes are attacked, deleted, or destroyed. The attacked nodes and their connections are removed, leading to fragmentation and loss of performance.

[0111] S52: Evaluation of network effectiveness under attack scenarios. The attack strategy consists of random attacks and five intentional attack strategies based on node weighted degree, betweenness centrality, closeness centrality, and K-core, covering node importance evaluation methods based on neighbor nodes, paths, and node positions, which can simulate different types of high-risk scenarios. By simulating the network operation under a single attack strategy, the network effectiveness represented by the remaining organizational network is evaluated to represent the organizational resilience of the emergency response network. The simulation results of the organizational network in the four stages of pre-disaster plan, emergency rescue, recovery and reconstruction, and post-disaster plan are shown in the figure below. Figure 4 and Figure 5 shown.

[0112] Random attack is a non-targeted attack method. It does not consider the difference in the importance of nodes in the network, but randomly selects nodes to attack. The intentional attack strategy based on node weighted degree focuses on attacking nodes with high connectivity or high edge weight in the network. These nodes usually play the role of "hubs" in the network. Betweenness centrality measures the importance of a node as a "bridge" in the network. Nodes with high betweenness centrality are necessary nodes on many shortest paths. Closeness centrality reflects the average shortest distance from a node to all other nodes in the network. A node with high closeness centrality means that it can quickly exchange information with other nodes. K-core is a network structure analysis method that divides the nodes in the network according to their K-core level. Nodes in high-level K-cores tend to have closer connections and higher importance in the network. By simulating the network operation under a single attack strategy, the network effectiveness reflected by the remaining organizational network is used to evaluate the organizational resilience of the emergency response network.

[0113] S53: Network attack simulation results Figure 6 and Figure 7 This indicates that different attack strategies have varying impacts on network performance. To explore the effectiveness of different attack strategies on network performance, we ranked the impact of different attack strategies on network performance in descending order, based on the order in which they reduced the network performance index to zero, indicating complete network collapse. The results are shown in Table 1. Deliberate attack strategies based on different node importance indices are simplistically represented using these indices. For example, "deliberate attack based on betweenness centrality" is simply written as "betweenness centrality" in the table.

[0114] Table 1 Ranking of the influence of different attack strategies on network effectiveness

[0115]

[0116] S54: Analysis of the strategic influence ranking results for organizational networks at each stage indicates that, in risk scenarios, the K-core-based node importance evaluation method is more effective and efficient for identifying "core nodes" that play a significant role in organizational networks. The K-core indicator prioritizes the position of nodes in the network for hierarchical classification, arguing that nodes located within the core of the network have greater influence. Looking at the strategic influence ranking results for organizational networks at each stage, the K-core-based deliberate attack strategy has the greatest destructive power on the network, indicating that, in risk scenarios, the K-core-based node importance evaluation method is more effective and efficient for identifying "core nodes" that play a significant role in organizational networks. The K-core indicator prioritizes the position of nodes in the network for hierarchical classification, arguing that nodes located within the core of the network have greater influence. This result reflects the core-periphery structure of emergency response organizational networks.

[0117] S55: As mentioned above, the performance of the actual response organization network varies significantly under different attack strategies, especially for the emergency rescue organization network. Table 1 shows that the K-core-based deliberate attack strategy has the greatest impact on the performance of this network. Therefore, the K-core layer in the emergency rescue network is listed.

[0118] S56: The organization that collapsed first under the K-core attack strategy indicates that organizations with the highest K-core values ​​are the most important nodes for network performance. Based on actual circumstances, an analysis of actual emergency response organizations demonstrates their crucial role in emergency response. The emergency rescue organization network collapsed first under the K-core attack strategy, indicating that organizations with the highest K-core values ​​are the most important nodes for network performance.

[0119] In step 6, the complex network-based emergency response organizational resilience evaluation method is used to couple multiple network attack strategies using the entropy weight method. The network effectiveness of the remaining organizational network is evaluated by simulating different high-risk scenarios to characterize the organizational resilience of the emergency response network. Specifically, the following are included:

[0120] S61: Using the entropy weight method to couple multiple network attack strategies, by simulating different high-risk scenarios, and after standardizing the data, the weight of each attack strategy is obtained by calculating the information entropy. The entropy weight method formula is as follows:

[0121]

[0122] Where: X ij is the standardized network performance index data, T ij is the normalized index value, e i is the information entropy calculation value, w i is the weight of each attack strategy obtained by the entropy weight method.

[0123] The entropy weight method is used to assign weights and calculate the weights of node attack strategies under various network effectiveness indicators, as shown in Table 2 (retain two decimal places). Among them, the intentional attack strategies based on different node importance indicators are concisely represented by this indicator.

[0124] Table 2 Weight calculation results for different node attack strategies

[0125]

[0126] The emergency response organization resilience evaluation indicators based on network effectiveness evaluation in step 6 are as follows:

[0127]

[0128] Where: F j represents the network performance evaluation value, w i is the attack strategy weight, f i is the performance index value; R represents the tissue toughness evaluation value, h i Use + / - according to the positive or negative impact of performance on toughness.

[0129] In step 7, the emergency response organizational resilience evaluation method based on complex networks is used to evaluate the organizational resilience of the emergency response network by simulating different high-risk scenarios. The network effectiveness of the remaining organizational network is used to characterize the organizational resilience of the emergency response network. Specifically, the following are performed:

[0130] S71: Using the entropy weight method to couple multiple network attack strategies, by simulating different high-risk scenarios, and after standardizing the data, the weight of each attack strategy is calculated by calculating the information entropy;

[0131] S72: Since the robustness, transmission, and recovery of organizational networks are positively correlated with organizational resilience, and the clustering is negatively correlated with organizational resilience, based on the network effectiveness evaluation results and their correlation with organizational resilience, the organizational resilience evaluation is achieved by simply averaging the network effectiveness evaluation results. That is, organizational resilience is a comprehensive reflection of the four network effectiveness.

[0132] S73: The results of organizational network effectiveness evaluation and organizational resilience evaluation are as follows: Figure 8 、 Figure 9 The toughness values ​​of some tissues are shown in Table 3.

[0133] Table 3 Tissue toughness evaluation results

[0134]

[0135] S74: The results in Table 3 show that: First, (1) from the perspective of organizational network effectiveness evaluation results, when the ratio of removed nodes is 0.6, the resilience and agglomeration of the pre-disaster emergency plan organizational network drop significantly. The reason is that before and after this, the intentional attack strategy based on proximity centrality successively removed the "provincial flood control headquarters" and "municipal flood control headquarters" nodes, causing highly harmful targeted damage to the network. Before this, the effectiveness of the organizational network at each stage was roughly as follows: the performance of the post-disaster emergency plan network was significantly higher, and the performance of the recovery and reconstruction network was significantly lower. The pre-disaster emergency plan network was slightly better than the emergency rescue network in terms of robustness and transmission. (2) from the perspective of organizational resilience evaluation results, when the ratio of removed nodes was 0.5, the organizational resilience of the emergency rescue network was 0.0305, and the organizational resilience of the recovery and reconstruction network was 0.0220, indicating that the organizational resilience of the actual response network tends to 0 at this time, and the organization has almost no response capability to deal with risks. Before this, the organizational resilience of the organizational network at each stage was ranked as follows: post-disaster plan > pre-disaster plan > emergency rescue > recovery and reconstruction. (3) Overall, the performance of each organizational network is consistent with organizational resilience. That is, the performance and resilience of the planned response organizational network based on the emergency plan are better than the actual response organizational network. The post-disaster plan is better than the pre-disaster plan, and the emergency rescue stage is better than the recovery and reconstruction stage.

[0136] S75: The organizational network formed by the emergency plan revised after the disaster performed best in terms of network performance and overall resilience, representing the optimal inter-organizational collaborative cooperation effect, indicating that the target effectiveness of the revised planned emergency response has been improved.

[0137] S76: The results of the organizational resilience evaluation show that although the effectiveness and resilience of the pre-disaster emergency plan organizational network are not as good as those of the revised post-disaster emergency plan, the overall performance is still better than the actual response network. This shows that even if it represents a lower inter-organizational collaborative response effectiveness target, there is still a gap between the actual emergency response and it. Therefore, improving organizational resilience and organizational coordination effects in the actual response process is a core need of current emergency management.

[0138] Based on any of the above embodiments or a combination of multiple embodiments, according to another aspect of the present invention, a complex network-based emergency response organization resilience evaluation system is provided, comprising:

[0139] The first main control module is used to build an organizational relationship matrix based on actual response data and emergency plan data to establish the relationship and interaction mechanism between various subjects in the emergency response process;

[0140] The second main control module is used to build an organizational network for the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix;

[0141] The third main control module is used to evaluate the importance of nodes using the node importance metrics of weighted degree, betweenness centrality, closeness centrality, and K-core;

[0142] The fourth main control module is used to evaluate the resilience of the network organization that represents the emergency response by simulating the network operation under a single attack strategy;

[0143] The fifth main control module is used to couple multiple network attack strategies using the entropy weight method, and then simulate different high-risk scenarios to obtain the weight of each attack strategy by calculating the information entropy;

[0144] The sixth main control module is used to evaluate the network effectiveness under various attack scenarios and derive the overall resilience level of the emergency response organization at each stage.

[0145] In addition, the simulation part of the system is mainly implemented based on Gephi software and Python's networkX module.

[0146] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A complex network-based emergency response organization resilience evaluation method, characterized by: The following steps are involved: Step 1: Based on the actual response data and emergency plan data, an organizational relationship matrix is ​​constructed to establish the relationship and interaction mechanism between various subjects in the emergency response process; Step 2: Construct an organizational network for the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix; Step 3: Use the weighted degree, betweenness centrality, closeness centrality, and K-core node importance metrics to evaluate node importance; Step 4: By simulating the network operation under a single attack strategy, the resilience of the network organization representing the emergency response is evaluated; Step 5: Use the entropy weight method to couple multiple network attack strategies, simulate different high-risk scenarios, and calculate the weight of each attack strategy by calculating the information entropy; Step six: Evaluate the network effectiveness under various attack scenarios to determine the overall resilience level of the emergency response organization at each stage.

2. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: Step 1 includes the following steps: (11) According to different stages of emergency response, a multi-source mixed data collection method is used to collect actual response data and emergency plan data; (12) Integrate, clean and pre-process the text data of the actual response data and emergency plan data obtained; (13) The organizational network is constructed through the relationship extraction method based on text analysis, and the information extraction and relationship expression between organizations are realized; (14) The co-occurrence frequency of organizations is used as the edge weight of organizational relationships to obtain the organizational relationship matrix, and the relationship and interaction mechanism between various subjects in the emergency response process are established.

3. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: Step 2 includes the following steps: (21) Construct the organizational network of the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix, and conduct network visualization; (22) Perform community division visualization, node affiliation is divided by node color, and node size is described according to weighted degree; (23) Through the topological structure of the organizational network, visualization of its components and measurement analysis of its overall structural indicators, we can find the structural characteristics and stage differences of the organizational response network.

4. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: Step three includes the following steps: (31) The importance of nodes is evaluated using weighted degrees. The greater the weighted degree, the greater the influence of the node and the more important the node. (32) Betweenness centrality is used to evaluate the importance of nodes. The higher the betweenness centrality, the more important the node is. (33) Closeness centrality is used to evaluate the importance of nodes. The greater the closeness centrality of a node, the more central the node is in the network and the closer it is to other nodes in the network. Preferably, the calculation model of the weighted degree includes: Where k i is the weighted degree of node i, h ij is the edge weight between node i and its neighbor node j; Preferably, the calculation model of betweenness centrality includes: Where BC(i) is the betweenness centrality of node i, V is the set of nodes in the network, and n st is the number of shortest paths between nodes s and t, n st (i) is the number of shortest paths between points s and t that pass through node i; Preferably, the calculation model of the closeness centrality includes: Where CC(i) is the closeness centrality of node i, N is the total number of nodes, and d ij is the shortest distance between nodes i and j.

5. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: Step 4 includes the following steps: (41) By systematically evaluating the importance of nodes, we can simulate network behavior under different attack strategies; (42) By simulating the network operation under a single attack strategy, the network effectiveness reflected by the remaining organizational network is evaluated to represent the organizational resilience of the emergency response network; (43) According to the attack strategy, the network performance index value is reduced to 0, that is, the order of network complete collapse is analyzed, and the influence of different attack strategies on network performance is sorted in descending order.

6. The complex network-based emergency response organization resilience evaluation method according to claim 5 is characterized in that: The simulation of network behaviors under different attack strategies includes: Simulate scenarios where nodes are attacked, deleted, or destroyed, removing the attacked nodes and their connections, leading to fragmentation and performance loss of the organization's network.

7. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: In step 5, the calculation formula of the entropy weight method includes: Where, X ij is the standardized network performance index data, T ij is the normalized index value, e i is the information entropy calculation value, w i is the weight of each attack strategy obtained by the entropy weight method.

8. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: Step five also includes: (51) The entropy weight method is used to couple multiple network attack strategies and simulate different high-risk scenarios. After standardizing the data, the weight of each attack strategy is obtained by calculating the information entropy. (52) Since the robustness, transmission and recovery of organizational networks are positively correlated with organizational resilience, and the clustering is negatively correlated with organizational resilience, based on the network effectiveness evaluation results and their correlation with organizational resilience, the organizational resilience evaluation is achieved by simply averaging the network effectiveness evaluation results.

9. The complex network-based emergency response organization resilience evaluation method according to claim 1 is characterized in that: In step 6, the emergency response organization resilience evaluation indicators based on network effectiveness evaluation are as follows: Where, F j represents the network performance evaluation value, w i is the attack strategy weight, f i is the performance index value; R represents the tissue toughness evaluation value, h i Use + / - according to the positive or negative impact of performance on toughness.

10. A complex network-based emergency response organization resilience evaluation system, characterized by: include: The first main control module is used to build an organizational relationship matrix based on actual response data and emergency plan data to establish the relationship and interaction mechanism between various subjects in the emergency response process; The second main control module is used to build an organizational network for the four stages of pre-disaster planning, emergency rescue, recovery and reconstruction, and post-disaster planning based on the organizational relationship matrix; The third main control module is used to evaluate the importance of nodes using the node importance metrics of weighted degree, betweenness centrality, closeness centrality, and K-core; The fourth main control module is used to evaluate the resilience of the network organization that represents the emergency response by simulating the network operation under a single attack strategy; The fifth main control module is used to couple multiple network attack strategies using the entropy weight method, and then simulate different high-risk scenarios to obtain the weight of each attack strategy by calculating the information entropy; The sixth main control module is used to evaluate the network effectiveness under various attack scenarios and derive the overall resilience level of the emergency response organization at each stage.

Images