System toughness key element mining method, equipment, medium and product

Abstract

The invention discloses a mining method and device for key elements of system toughness, a medium and a product, and relates to the technical field of system toughness evaluation.The method comprises the steps that a system super-network model is constructed according to a target system and a typical application scene of the system, and the target system is an emergency rescue unmanned aerial vehicle system; the system super network model comprises a task layer, a collaboration layer and a physical layer, the task layer is the top layer of the target system architecture, the collaboration layer is the middle layer of the target system architecture, and the physical layer is the entity basis of the target system architecture; determining a system toughness evaluation index according to the system hypernetwork model; constructing a first system toughness evaluation model based on the system super-network model and the system toughness evaluation indexes; and based on a system evaluation key scene in the typical application scene of the system, and in combination with the first system toughness evaluation model, carrying out mining to obtain system toughness key elements. According to the method, the key elements of the system toughness can be excavated, and the corresponding system equipment is optimized, so that the system toughness level is improved.

Description

Technical Field

[0001] The present application relates to the technical field of system resilience assessment, and in particular to a method, equipment, medium and product for mining key elements of system resilience. Background Art

[0002] A system consists of a series of interacting systems or elements that work together to achieve specific functions that a single system cannot accomplish independently, such as equipment systems, concept systems, search and rescue systems, etc. Resilience refers to the ability of a system to maintain its functional stability when resisting interference and after being disturbed. In a complex and changing environment, emergency rescue drone systems often encounter various types of interference such as random failures and climate change. Especially under extreme conditions, complex situations such as large-scale functional failures may occur. It can be seen that the resilience of the system is crucial to its ability to perform tasks. At present, how to improve the resilience level of the system is an urgent problem to be solved. Summary of the invention

[0003] The purpose of this application is to provide a method, equipment, medium and product for mining key elements of system resilience, which can identify key elements that affect the level of system resilience, thereby improving the level of system resilience by focusing on the performance of key elements of system resilience.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a method for mining key elements of system resilience, including:

[0006] According to the acquired target system and the typical application scenarios of the system, a system hyper-network model is constructed; the target system is an emergency rescue drone system, and the system hyper-network model includes a task layer, a collaboration layer and a physical layer, wherein the task layer is the top layer of the target system architecture, the collaboration layer is the middle layer of the target system architecture, and the physical layer is the entity foundation of the target system architecture;

[0007] Determining a system resilience evaluation index according to the system hypernetwork model;

[0008] Based on the system hypernetwork model and the system resilience evaluation index, construct a first system resilience evaluation model;

[0009] Based on the first system resilience assessment model and system assessment key scenarios, key elements of system resilience are mined; the system assessment key scenarios are determined based on typical application scenarios of the system.

[0010] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for mining key elements of system resilience as described above.

[0011] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for mining key elements of system resilience described above.

[0012] In a fourth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method for mining key elements of system resilience described above.

[0013] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0014] The present application provides a method, equipment, medium and product for mining key elements of system resilience. According to the target system and the typical application scenarios of the system, a system hyper-network model is constructed. The target system is an emergency rescue drone system. The system hyper-network model includes a task layer, a collaboration layer and a physical layer. The task layer is the top layer of the target system architecture, the collaboration layer is the middle layer of the target system architecture, and the physical layer is the entity foundation of the target system architecture. According to the system hyper-network model, the system resilience evaluation index is determined. Based on the system hyper-network model and the system resilience evaluation index, a first system resilience evaluation model is constructed. Based on the system evaluation key scenarios in the typical application scenarios of the system, and in combination with the first system resilience evaluation model, the key elements of system resilience are mined. The present application identifies the key elements of system resilience in different key scenarios, and pays attention to and optimizes the system equipment corresponding to the key elements in different scenarios, so that the system resilience level can be improved by optimizing the performance of the key elements of system resilience, and the stability and fault tolerance of the system can be enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 This is an application environment diagram of a method for mining key elements of resilience of a system in one embodiment of the present application;

[0017] Figure 2A flowchart of a method for mining key elements of system resilience provided in one embodiment of the present application;

[0018] Figure 3 for Figure 2 A detailed flow chart of step 201;

[0019] Figure 4 for Figure 2 A detailed flow chart of step 202;

[0020] Figure 5 for Figure 2 A detailed flow chart of step 203;

[0021] Figure 6 for Figure 2 A detailed flow chart of step 204;

[0022] Figure 7 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0025] The method for mining key elements of system resilience provided in the embodiments of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the target system and the typical application scenarios of the system to the server 104. After the server 104 receives the target system and the typical application scenarios of the system, the server 104 constructs a system hypernetwork model according to the target system and the typical application scenarios of the system; determines the system resilience evaluation index according to the system hypernetwork model; constructs a first system resilience evaluation model based on the system hypernetwork model and the system resilience evaluation index; based on the first system resilience evaluation model and the system evaluation key scenarios in the system typical application scenarios, the key elements of system resilience are mined. The server 104 can feedback the obtained key elements of system resilience to the terminal 102. In addition, in some embodiments, the method for mining key elements of system resilience can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly process the target system and typical application scenarios of the system, or the server 104 can obtain the target system and typical application scenarios of the system from the data storage system and process them.

[0026] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, and IoT devices, and the server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0027] In an exemplary embodiment, Figure 2 As shown, a method for mining key elements of system resilience is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 201 to 204. Among them:

[0028] Step 201, construct a system hypernetwork model according to the target system and typical application scenarios of the system; the target system is an emergency rescue drone system, and the system hypernetwork model includes a task layer, a collaboration layer and a physical layer. The task layer is the top layer of the target system architecture, the collaboration layer is the middle layer of the target system architecture, and the physical layer is the entity foundation of the target system architecture.

[0029] Step 202: Determine a system resilience evaluation index based on the system hypernetwork model.

[0030] Step 203: construct a first system resilience assessment model based on the system hypernetwork model and the system resilience assessment index.

[0031] Step 204, based on the first system resilience assessment model and the system assessment key scenarios, key elements of system resilience are mined; the system assessment key scenarios are determined based on typical application scenarios of the system.

[0032] By implementing the above-mentioned steps 201 to 204, the present application identifies the key elements of system resilience in different key scenarios, and pays attention to and optimizes the system equipment corresponding to the key elements in different scenarios, thereby improving the system resilience level and enhancing the stability and fault tolerance of the system by optimizing the performance of the key elements of system resilience.

[0033] In another exemplary embodiment of the present application, Figure 3 As shown, step 201 involves constructing a system hypernetwork model according to the target system in detail; wherein a complex system can be abstracted into a graph G consisting of a point set V and an edge set E, that is: G = (V, E). The abstracted system hypernetwork model can clearly reflect the calling relationship between the nodes of the unmanned system and facilitate the reliability analysis of the unmanned system. Depending on the different calling relationships of the emergency rescue drone system, the graph G can be a directed graph or an undirected graph, and the storage form of the graph G can be a matrix or a linked list. Based on this, by analyzing the system to be evaluated (i.e., the target system), the topological structure between its nodes and the calling relationship between the nodes are discovered, and a system hypernetwork model is constructed, specifically including:

[0034] Step 211, determine the system components according to the target system and the typical application scenarios of the system. The system components include: the structure, function, capability, input, output and other properties of the system. Taking the emergency rescue drone system as an example, its elements include: perception, recognition, decision-making, execution, evaluation, diagnosis, etc. The specific implementation process of this step is:

[0035] According to the system application scenarios, construct typical application scenarios of the system, analyze the typical operation process of the system, analyze the system structure composition, functional composition, operation process and the interaction mode and process of each component during the operation process, and extract the set N of system components.

[0036] Taking the emergency rescue drone system as an example, its typical application scenario is the rescue of personnel in a complex fire environment. Considering the special situations that may occur at any time in the rescue scenario, the system performance will be disturbed to a certain extent. Therefore, in this scenario, the task of the emergency rescue drone system is to search the fire scene, analyze the dangerous situation, find the objects that need rescue and organize drone units for rescue, evaluate the damage of the emergency rescue drone system caused by the sudden situation of the fire scene and conduct command and dispatch. From this analysis, the emergency rescue drone system has six capabilities: perception, recognition, decision-making, execution, evaluation, and diagnosis, which are also the ability elements of the system. The command structure of the emergency rescue drone system is divided into commanders and executors, and its system architecture is divided into command and action, that is, the structural elements of the system. Therefore, the system component element set N contains a total of 8 system elements: command, action, perception, recognition, decision-making, execution, evaluation, and diagnosis.

[0037] Step 2012, for typical application scenarios of the system, based on the system components, using the DoDAF system modeling method, construct an architecture-function-task model, including the capability view (CV-2) of the system tasks, the structural view (SV-4) of the architecture, and the functional view (OV-5b) of the system functions.

[0038] Step 213, converting the architecture-function-task model into a system hypernetwork model. The specific implementation process of this step is:

[0039] Based on the structure-function-task model, the key elements at different levels of the system and the relationship between the key elements are explored from different perspectives, and then a three-layer logical architecture is extracted: the task layer, the collaboration layer, and the physical layer. The intra-layer relationship and inter-layer coupling relationship of each logical architecture are analyzed to establish a system super network model. Among them, the task layer is the top layer of the system architecture; the collaboration layer is the middle layer of the system architecture; and the physical layer is the physical foundation of the system architecture.

[0040] Taking the emergency rescue drone system as an example, its task layer contains six types of task decision elements: perception, recognition, decision-making, execution, evaluation, and diagnosis. It is responsible for the construction and driving of rescue tasks, as well as the generation of tasks for dangerous situations. The collaborative layer is mainly composed of information elements of commanders and actors, which controls the collection and sharing of dangerous situation information and realizes the processing of dangerous situation tasks through the rapid flow of information. The physical layer is composed of specific drone equipment and facilities. Each entity unit operates autonomously according to the decision rules or accepts and executes superior instructions. The driving of the central task decision link in the task layer needs to rely on the transmission processing of the information interaction architecture, and the transmission processing of the information interaction architecture in the collaborative layer needs to rely on the normal operation of the equipment entities and communication networks in the physical layer.

[0041] In another exemplary embodiment of the present application, Figure 4As shown, the step 202 involves determining the system resilience evaluation index according to the system hypernetwork model, which specifically includes:

[0042] Step 2021, based on the system hypernetwork model, determine the system task performance indicators and system network structure indicators; the system task performance indicators and system network structure indicators are the analysis results obtained after analyzing the system hypernetwork model based on the system's task conditions, performance status and structural status, and provide a quantitative means for the constantly changing status during the system evolution process.

[0043] Step 2022, based on the typical application scenarios of the system, simulate the system operation stage, system confrontation stage and system reconstruction stage, analyze the exchange mode between system elements in different stages of system resilience, and calculate the indicator change process of system task performance indicators and system network structure indicators in the system operation stage, system confrontation stage and system reconstruction stage. The spatiotemporal evolution process of system resilience in the three stages of operation, confrontation and reconstruction is summarized through the indicator change process.

[0044] Taking the emergency rescue drone system as an example, the system is aimed at emergency rescue in high-risk emergencies such as earthquake relief and fire rescue. Its system mission performance indicators include but are not limited to rescue mission closure efficiency, average rescue mission closure time, etc., and the system network structure indicators include but are not limited to network average degree, network average betweenness, etc. Modeling and simulation of the system operation, confrontation and reconstruction stages for the fire rescue scenario can calculate the change process of the system mission performance indicators and system network structure indicators. Simulate the resilient system and the non-resilient system respectively. Resilience in the example refers to the ability to evaluate the current state of the system and reallocate system resources.

[0045] Step 2023, based on the spatiotemporal evolution characteristics of system resilience, determine the system resilience evaluation index to characterize the system capability. The specific implementation process of this step is as follows:

[0046] Design system robustness indicators. Design system robustness indicators according to the spatiotemporal evolution characteristics of resilience during the operation phase of the system. Optionally, but not limited to, design system robustness level indicators. The system robustness level indicator L characterizes the system's ability to absorb disturbances. It is defined as the ratio of the system capability curve area under disturbance application conditions to the system capability curve area under no disturbance conditions, provided that the system maintains above the normal operating performance threshold. The expression is:

[0047]

[0048] Among them, MC(t) is the system capacity value at time t; MC(t 0 ) is the system capacity value at the initial state; It indicates the cumulative capacity of the system to maintain normal working performance after being attacked; It refers to the cumulative ability of the system to maintain normal working performance when it is not attacked.

[0049] Design system resistance index. Design system resistance index according to the spatiotemporal evolution characteristics of the system's resilience in the confrontation stage. Optionally, but not limited to, design system resistance level index. The system resistance level index D measures the system's ability to absorb disturbances and resist system performance degradation. It can be expressed as a negative exponential function of the degree of capacity reduction d and the efficiency of capacity reduction time, and the expression is:

[0050]

[0051] Among them, d is the degree of capacity reduction, which refers to the degree of deviation between the system capacity reduced to the minimum performance level after the disturbance and the normal operating state capacity, and the expression is: MC(t d ) is the value when the system capacity is reduced to the lowest state; v d To reduce the time efficiency, the expression is e is an exponential function.

[0052] Design system recovery indicators. Design system recovery indicators based on the spatiotemporal evolution characteristics of resilience in the reconstruction phase of the system. Optionally, but not limited to, design system recovery level indicators. The system recovery level indicator R refers to the ability of the system to complete tasks after recovery strategies such as node backup and node replacement. R can be expressed as the system task capability recovery degree r and capability recovery time efficiency v. r The exponential function of

[0053]

[0054] Among them, r is the degree of ability recovery, expressed as MC(t r ) is the value when the system capacity is reduced to the lowest state, MC(t s ) is the value after the system capacity is restored; v r The efficiency of ability recovery time is expressed as

[0055] Design system indicators for the entire process. Design system resilience indicators based on the spatiotemporal evolution characteristics of resilience throughout the entire process of the system. Optionally, but not limited to, design system resilience triangle area indicators. The dynamic changes in the height of performance degradation during standard operation and the anti-reconstruction phase of the system resilience triangle area can be expressed by the ratio of the difference area of ​​the system capability curve under undisturbed and disturbed conditions to the area of ​​the system capability curve under undisturbed conditions, and the expression is:

[0056]

[0057] in, It represents the difference between the cumulative working capacity of the system under undisturbed and disturbed conditions. It indicates the cumulative working capacity of the system under undisturbed conditions.

[0058] In another exemplary embodiment of the present application, Figure 5 As shown, the construction of the first system resilience assessment model based on the system hypernetwork model and system resilience assessment indicators involved in step 203 is introduced in detail. Based on the system resilience assessment indicators, a system resilience assessment model is constructed around two key issues: how to directly evaluate the system resilience based on the system components and how to identify the key factors affecting the system resilience. Based on the disturbance injection strategy, the system under different conditions is simulated and tested, and the resilience of the simulation data is calculated according to the constructed system resilience assessment indicators to obtain the system element status and the corresponding system resilience level as the input of the system resilience assessment model, and the system resilience assessment model performance indicators are analyzed based on the model input. Specifically including:

[0059] Step 2031, based on the system components in the system hypernetwork model, construct a first system disturbance strategy; the first system disturbance strategy is multiple. The specific implementation process of this step is as follows:

[0060] First, the disturbance strategy is designed around the system components. Then, the initial disturbance is injected at the beginning of the system simulation. The method of injecting disturbance is to cause the system elements to fail, and the failure modes of different system elements are different.

[0061] Taking the emergency rescue drone system as an example, the system components include perception, recognition, decision-making, execution, evaluation, diagnosis, command and action. The above 8 system elements can be set to 2 initial states, normal state (1) and fault state (0). Through the different state combinations of different system elements, a total of 2 8 = 256 perturbation strategies to form a perturbation space.

[0062] Step 2032: Generate system disturbance data according to the system disturbance strategy, and construct a system resilience assessment data set (i.e., a first sample data set). The specific implementation process of this step is as follows:

[0063] Each first system perturbation strategy is input into the target system for simulation, and the target system resilience under each first system perturbation strategy is calculated based on the system resilience evaluation index, and the calculation result of the system resilience index corresponding to each first system perturbation strategy is obtained; based on the calculation result of the system resilience index, a clustering algorithm is used to cluster the discretized simulation data, and the data is divided into n categories as system resilience labels, and the system resilience label corresponding to each first system perturbation strategy is obtained; the system resilience label is high resilience or low resilience; each sample data in the first sample data set includes a first system perturbation strategy and its corresponding system resilience label.

[0064] Taking the emergency rescue drone system as an example, the disturbance data is an 8-dimensional vector generated by the disturbance strategy, and each dimension has two states, "0" and "1", representing the element "not faulty" and "faulty" respectively. The model uses two outputs, "high resilience" and "low resilience", to represent the resilience level of the system.

[0065] Step 2033, construct an initial system resilience assessment model; the initial system resilience assessment model is constructed based on a machine learning algorithm, and the machine learning algorithm may be selected but is not limited to: logistic regression, decision tree, support vector machine, KNN algorithm, random forest, etc.

[0066] Step 2034, divide the first sample data set into a first training data set and a first test data set, use the first training data set to train the initial system resilience assessment model to obtain the first system resilience assessment model, and use the first test data set to test the first system resilience assessment model to calculate the accuracy of the first system resilience assessment model. The specific implementation process of this step is as follows:

[0067] First, the performance level of the system components (i.e., the perturbation strategy) is taken as the feature vector X, and the system resilience level (i.e., the system resilience label) is taken as the data label Y. Then, the first sample data set is divided, and 70% of the perturbation strategies in the first sample data set are taken as the first training data set, and 30% of the perturbation strategies are taken as the first test data set. On this basis, the initial system resilience assessment model is trained using the first training data set to obtain the first system resilience assessment model, and the first test data set is used to test the first system resilience assessment model, and the accuracy of the first system resilience assessment model is calculated to evaluate the training effect.

[0068] In another exemplary embodiment of the present application, Figure 6As shown, the key elements of system resilience are mined based on the first system resilience assessment model and system assessment key scenarios involved in step 204. Taking into account the dynamic evolution characteristics of the system, such as multiple actual operating states and large fluctuations, not all system elements can be associated with system resilience. The high resilience and low resilience states of the system may be very different in specific system elements at certain levels. Only when the performance characteristics of the key elements of the system show large differences, the prediction of the system resilience state based on the corresponding characteristic indicators has higher accuracy. Therefore, the key elements of system resilience that distinguish the high resilience and low resilience states of the system can be identified. Specifically including:

[0069] Step 2041, determine the key scenarios for system evaluation. Design the key scenarios for system evaluation based on the task completion process of the typical application scenarios of the system, including system scenario elements and system component elements.

[0070] Taking the emergency rescue drone system as an example, its typical application scenarios include but are not limited to earthquake relief, fire rescue, etc. Taking fire rescue as an example, the system components in the key scenarios of the system evaluation include but are not limited to perception, recognition, decision-making, execution, evaluation, diagnosis, command, action and other elements; the system scenario elements in the key scenarios of the system evaluation include but are not limited to the random damage of drone equipment caused by unstable fire outbreaks, and the limited perception ability of drone equipment caused by smoke.

[0071] Step 2042, based on the system assessment key scenario, screen out system disturbance elements from the system components, and construct a second system disturbance strategy based on the system disturbance elements, and there are multiple second system disturbance strategies.

[0072] Taking the emergency rescue drone system in the fire rescue scenario as an example, due to the scene element (i.e., system disturbance element) in which the drone equipment will be randomly damaged in the fire scene, the information architecture such as command and action will fail, so the disturbance strategy of this element can be set to "1"; due to the scene element in which the drone equipment's perception ability is limited due to smoke in this scene, tasks such as perception and recognition will fail, so the disturbance strategy of this element can be set to "1". Therefore, the construction of the second system disturbance strategy is completed based on the system disturbance element.

[0073] Step 2043, based on the first system resilience assessment model and the second system disturbance strategy, calculate the contribution rate of each system element in the system components, and sort them according to the contribution rate of each system element to determine the key elements of system resilience. The specific implementation process of this step is as follows:

[0074] Remove the current element from the system components; construct a second sample data set, and divide the second sample data set into a second training data set and a second test data set, each sample data in the second sample data set includes a second system perturbation strategy and its corresponding system resilience label; use the second training data set to train the first system resilience assessment model to obtain an updated first system resilience assessment model, and use the second test data set to test the updated first system resilience assessment model, and calculate the accuracy Acc(-X) of the updated first system resilience assessment model; according to the accuracy of the updated first system resilience assessment model and the accuracy of the first system resilience assessment model, calculate the accuracy drop value Acc-Acc(-X) after removing the current element to obtain the contribution rate of the current element; the greater the accuracy drop, the greater the role played by the removed system element in the prediction of the system resilience state, that is, the greater the contribution rate; take each system element in the system components as the current element one by one, and repeat the above steps to obtain the contribution rate of each system element; sort according to the contribution rate of each system element, and determine the system elements with the highest ranking as the key elements of system resilience.

[0075] Taking the emergency rescue drone system as an example, its typical application scenario is the rescue of personnel in complex fire environments. The system takes emergency rescue in complex rescue scenarios such as earthquakes and fires as its application target, takes the static map of the scene range and the rescue mission target as input, and monitors and analyzes the dangers of the entire disaster relief scene by coordinating and commanding various types of drones including information collection, command decision-making, and rescue implementation; calculates and outputs the deployment plan of drone resources, handles various emergencies at any time, and completes emergency rescue missions.

[0076] According to the typical application scenarios of the above system, the typical application scenarios of emergency rescue drones are set up based on the system simulation environment as follows:

[0077] Taking complex scene rescue under extreme conditions as the starting point, a task scenario of drone swarm handling dangerous targets is constructed. In this scenario, drone units are divided into emergency command drones, search and rescue drones, etc. The emergency command drone is responsible for periodically receiving and summarizing the dangerous situation collected by each search and rescue drone, evaluating the status of each dangerous situation, and deploying drones to perform search and rescue tasks through the auction algorithm drive; when the system is suddenly damaged, resources are allocated in time to restore the system capabilities. Search and rescue drones are divided into perception terminals and emergency terminals. The perception terminal is responsible for receiving dangerous information in the area to which the set belongs and transmitting it to command drones in various locations; the emergency terminal is equipped with rescue equipment and receives commands from command drones to deal with disasters. It should be noted that this method focuses on the resilience assessment of the system, and the actual equipment is simplified to particles with specific capabilities, retaining only parameters such as speed, position, direction, rescue radius and communication radius.

[0078] According to the above assumptions, a fire rescue scenario is constructed and the system is simulated. Assume that a chemical plant suddenly explodes, igniting nearby residential buildings and triggering a large-scale fire. It is necessary to organize drones to suppress the fire and search and rescue survivors and key equipment. Taking into account the sudden dangers such as deflagration and collapse that may occur at any time in the rescue scenario, the performance of the system will be affected to a certain extent, and the search and rescue drones may suffer losses. Therefore, in this scenario, the task of the drone system is to search the fire scene, analyze the danger, find the objects in need of rescue and organize drone units for rescue, evaluate the damage to the drone system caused by the sudden situation at the fire scene, and conduct command and dispatch. The specific settings of the scene are as follows:

[0079] The emergency rescue drone system in the scene includes 4 mutually communicating command drones and 56 search and rescue drones (24 emergency terminals and 32 sensing terminals), and communication rules are constructed to form a rescue system. The scene is modeled as a square area consisting of 1600 grid points, each of which is a unit area. The initial number of rescue targets and fire points to be extinguished is 400, scattered within the disaster area. When we carry out rescue, new fire points will continue to spread with a certain probability. The probability can be set to a 10% probability of producing a fire point for each fire location; when the simulation reaches the 400th time step, the assumed scene will explode, destroying 50% of the sensing terminals and emergency terminals, making them incapable.

[0080] The command drones perform complex global situation assessment, defect diagnosis and auction-style task decision-making. Each command drone complexly dispatches 6 emergency terminals and 8 perception terminals in an area. During the operation of the system, the command drone receives the disaster situation information collected by each perception terminal through the communication network at a certain frequency (every 50 simulation steps), generates a global situation, and evaluates the overall situation, mainly evaluating the system efficiency with the efficiency of fire extinguishing as the core. The specific execution is divided into two cases: 1) If the overall system efficiency meets the top-level task requirements at the current moment (reflected in the simulation as the efficiency is not lower than the normal working threshold), the task decision is allocated to the execution terminal in the control area of ​​the command drone based on the auction algorithm; 2) If the overall efficiency does not meet the top-level mission requirements at the current moment, the defect diagnosis environment is entered. The command drones communicate with each other on demand, cut off the area with the lowest efficiency as the allocated area, and further diagnose according to the perception efficiency and fire extinguishing efficiency of the allocated area to determine the standby perception terminal that needs to be allocated at the current moment. After determining the transfer area and transfer terminal type, they are sent in the form of instructions to the command terminal to which the transferred terminal belongs, which executes the transfer instructions to complete the system reconstruction.

[0081] The search and rescue drone is divided into a rescue terminal and a perception terminal. The rescue terminal is responsible for handling the disaster, and the perception terminal is responsible for collecting disaster information. Optionally but not limited to, the perception range of the perception terminal is set to 20 unit areas, the flight speed is 4 unit lengths per time step, and it can communicate with the command terminal of 100 unit lengths; the flight speed of the rescue terminal is 3 unit lengths per time step, and it can communicate with the command terminal of 100 unit lengths. The specific task execution process is as follows: 1) In the initial state, the perception terminal can perceive and identify disaster information within a certain range, and autonomously explore and perceive the area; if there is no unsurveyed area, the perception terminal requests the command drone to obtain the coordinates of other unsurveyed locations in the area through the communication network. After receiving the request, the command drone sends the corresponding coordinates to the perception terminal. After receiving the coordinates, the perception terminal moves to the target area (due to the uncertainty and dynamic time-varying characteristics of the disaster site, the perception task is time-sensitive, and the perceived area will become the area to be perceived again after a period of time, such as 25 simulation steps). 2) After receiving the order, the rescue terminal will execute the task of heading to the target location to handle the disaster. After maneuvering to the vicinity of the target location, it will use the onboard equipment to extinguish the fire. After completing the task, it will execute the next disaster handling task. When the search and rescue drone completes all tasks in the current area, it will switch to standby status.

[0082] Therefore, the simulation operation process of this system can be divided into two different situations:

[0083] (1) Normal operation stage: The system operates normally, and the three types of drones, namely perception, rescue, and command, interact with each other and form a task chain in a cycle.

[0084] 1) Perception, performed by the perception terminal, which detects whether there is an area to be explored within the field of view, thereby performing mobile perception.

[0085] 2) Identification: The sensing terminal identifies the dangerous situation within the field of view and uploads it to the command drone.

[0086] 3) Decision-making: Command the drone to store and evaluate the target information, drive the task decision-making by multi-round auction algorithm, intelligently match the actual target-execution, and send the target information to the corresponding rescue terminal.

[0087] 4) Execution: The emergency terminal handles the disaster according to the command information.

[0088] (2) Resilience reconstruction stage

[0089] After the system is damaged by an unexpected event, six processes including perception, identification, decision-making, execution, evaluation and diagnosis are carried out to restore the system's capabilities.

[0090] 1) Perception: This link is the same as the normal stage. The perception terminal performs perception and detects whether there are unexplored areas in the field of view through area detection, and then performs maneuvers.

[0091] 2) Identification: Same as the normal stage, the sensing terminal identifies the danger and uploads it.

[0092] 3) Contact, the decision-making stage is different from the normal stage. The command drone stores and evaluates the target information, and then makes a task decision through a multi-round auction algorithm to achieve the execution matching of target-execution and send the target information. The command drone will also transmit the target situation information to other command drones for global situation interaction and fusion.

[0093] 4) Execution: The emergency terminal handles the disaster according to the command information.

[0094] 5) Evaluation: During the evaluation phase, the perception effectiveness and rescue efficiency are evaluated, and then the communication effectiveness and task effectiveness are evaluated to obtain the overall capability evaluation result of the system.

[0095] 6) Diagnosis: The diagnosis phase locates defects through regional performance comparison, and then diagnoses the allocation of perception resources and emergency resources, communication network adjustments and task area division.

[0096] In the above decision-making process, the task target response based on the multi-round auction algorithm is as follows:

[0097] The auction algorithm solves the task allocation problem by simulating human auction activities. The allocated tasks are the commodities of the auction activities, and the subjects are the bidders of the auction activities. The task allocation is completed through bidding. The task allocation problem is to finally allocate each commodity (target) to the bidder (rescue terminal) and maximize the benefits as much as possible through the market price mechanism. The bidder's price valuation is determined by the distance between the bidder and the target and the number of tasks the bidder has obtained. The closer the distance, the higher the valuation; the more tasks have been obtained, the lower the valuation.

[0098] At the beginning, all bidders (emergency terminals) are eligible to participate in the auction. All bidders who can participate in the auction and the goods to be allocated are auctioned. The parameters input into the auction algorithm also include the bidders' valuations of different goods. In each auction, some goods will be allocated to bidders. If a bidder is allocated more than three goods, he will be disqualified from bidding and will no longer be able to participate in subsequent auctions. This mechanism ensures that each entity does not take on too many tasks.

[0099] If all the goods have been allocated after the auction, the allocation is considered complete and the goods allocation result is output. If the goods have not been allocated after the auction, but there are no bidders with bidding qualifications, it means that the current subject is not enough to complete the existing task, and the goods allocation result is also output to complete the task allocation. If the goods have not been allocated after the auction and there are still bidders with bidding qualifications, the next auction will continue until the end condition is met. Since each auction will allocate goods, the algorithm can complete the task allocation number in a finite number of cycles.

[0100] According to the above process, a disaster simulation scenario is constructed to perform system simulation. The method for mining key elements of system resilience described in this application is adopted to construct a first system resilience assessment model based on the simulation results, and combined with the first system resilience assessment model, the key elements of system resilience are mined to provide guidance for the design and improvement of the resilience system.

[0101] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store processing data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for mining key elements of system resilience is implemented.

[0102] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0103] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0104] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0105] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0106] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0107] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0108] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0109] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for mining key elements of system resilience, characterized in that: The method for mining the key elements of system resilience includes: According to the acquired target system and the typical application scenarios of the system, a system hyper-network model is constructed; the target system is an emergency rescue drone system, and the system hyper-network model includes a task layer, a collaboration layer and a physical layer, wherein the task layer is the top layer of the target system architecture, the collaboration layer is the middle layer of the target system architecture, and the physical layer is the entity foundation of the target system architecture; Determining a system resilience evaluation index according to the system hypernetwork model; Based on the system hypernetwork model and the system resilience evaluation index, construct a first system resilience evaluation model; Based on the first system resilience assessment model and system assessment key scenarios, key elements of system resilience are mined; the system assessment key scenarios are determined based on typical application scenarios of the system.

2. The method for mining key elements of system resilience according to claim 1, characterized in that: The system hypernetwork model is constructed according to the acquired target system and the typical application scenarios of the system, specifically including: Determine the system components according to the target system and typical application scenarios of the system; According to the system components, a system architecture-function-task model is constructed using the DoDAF modeling method; the system architecture-function-task model includes a capability view of system tasks, a structural view of the system architecture, and a functional view of system functions; The architecture-function-task model is converted into the hierarchical hypernetwork model.

3. The method for mining key elements of system resilience according to claim 1, characterized in that: Determining the system resilience evaluation index according to the system hypernetwork model specifically includes: According to the system hypernetwork model, a system task performance index and a system network structure index are determined; the system task performance index and the system network structure index are analysis results obtained after analyzing the system hypernetwork model based on the system task situation, performance status and structure status; Based on the typical application scenarios of the system, the system operation stage, system confrontation stage and system reconstruction stage are simulated, and the change process of the system task performance index and the system network structure index in the system operation stage, system confrontation stage and system reconstruction stage is calculated to obtain the spatiotemporal evolution characteristics of the system resilience; According to the spatiotemporal evolution characteristics of the system resilience, an evaluation index of the system resilience is determined.

4. The method for mining key elements of system resilience according to claim 1, characterized in that: The system resilience evaluation index includes a system robustness level index, a system resistance level index, a system recovery level index and a system resilience triangle area index; The calculation process of the system robustness level index is: Among them, L is the system robustness index; MC(t) is the system capability value at time t; MC(t0) is the system capability value at the initial state; It indicates the cumulative capacity of the system to maintain normal working performance after being attacked; It indicates the cumulative ability of the system to maintain normal working performance when it is not attacked; The calculation process of the system resistance level index is: Among them, D is the system resistance level index; d is the degree of capacity reduction, v d is the ability to reduce time efficiency, e is an exponential function; The calculation process of the system recovery level index is: Among them, R is the system recovery level index; r is the degree of capacity recovery, v r Capacity recovery time efficiency; The calculation process of the toughness triangle area index of the system is: Among them, τ is the toughness triangle area index of the system; It represents the difference between the cumulative working capacity of the system under undisturbed and disturbed conditions. It indicates the cumulative working capacity of the system under undisturbed conditions.

5. The method for mining key elements of system resilience according to claim 1, characterized in that: The constructing a first system resilience assessment model based on the system hypernetwork model and the system resilience assessment index specifically includes: Based on the system components in the system hypernetwork model, construct a first system perturbation strategy; the first system perturbation strategy is multiple; Inputting each first system disturbance strategy into the target system for simulation, and calculating the target system resilience under each first system disturbance strategy based on the system resilience evaluation index, to obtain a system resilience index calculation result corresponding to each first system disturbance strategy; Based on the calculation result of the system resilience index, a clustering algorithm is used to obtain a system resilience label corresponding to each first system perturbation strategy; the system resilience label is high resilience or low resilience; Constructing a first sample data set and an initial system resilience assessment model, and dividing the first sample data set into a first training data set and a first test data set; each sample data in the first sample data set includes a first system perturbation strategy and its corresponding system resilience label, and the initial system resilience assessment model is constructed based on a machine learning algorithm; The initial system toughness assessment model is trained using the first training data set to obtain the first system toughness assessment model, and the first system toughness assessment model is tested using the first test data set to calculate the accuracy of the first system toughness assessment model.

6. The method for mining key elements of system resilience according to claim 5, characterized in that: Based on the first system resilience assessment model and the key system assessment scenarios, the key elements of system resilience are mined and obtained, specifically including: Based on the key scenario of the system evaluation, a system disturbance factor is obtained from the system components, and a second system disturbance strategy is constructed based on the system disturbance factor; the second system disturbance strategy is multiple; Based on the first system resilience assessment model and the second system disturbance strategy, the contribution rate of each system element in the system components is calculated, and the system elements are ranked according to the contribution rate to determine the key elements of system resilience.

7. The method for mining key elements of system resilience according to claim 6, characterized in that: The step of calculating the contribution rate of each of the system components based on the first system resilience assessment model and the second system disturbance strategy, and sorting the system components according to the contribution rate to determine the key elements of system resilience specifically includes: Removing the current element from the system components; Constructing a second sample data set, and dividing the second sample data set into a second training data set and a second test data set; each sample data in the second sample data set includes a second system perturbation strategy and its corresponding system resilience label; Using the second training data set to train the first system resilience assessment model to obtain an updated first system resilience assessment model, and using the second test data set to test the updated first system resilience assessment model to calculate the accuracy of the test of the updated first system resilience assessment model; Obtaining a contribution rate of a current factor according to an accuracy rate of a test performed on the updated first system resilience assessment model and an accuracy rate of the first system resilience assessment model; Taking each of the system components as the current element one by one, the above steps are executed cyclically to obtain the contribution rate of each system element; Each system element is ranked according to its contribution rate to determine the key elements of system resilience.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for mining key elements of system resilience as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements a method for mining key elements of system resilience as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, it implements a method for mining key elements of system resilience as described in any one of claims 1-7.

Images