Cross-domain unmanned cluster multi-stage comprehensive toughness evaluation method and device and computer equipment

Abstract

The invention relates to a cross-domain unmanned cluster multi-stage comprehensive toughness evaluation method and device and computer equipment, and the method comprises the steps: deeply analyzing a cross-domain unmanned cluster multi-layer coupling failure mechanism, and constructing a toughness process model covering four stages of prevention, degradation, recovery and reconstruction; and specific measurement methods of a corresponding preventability index, a robustness index, a restorability index and a reconfigurability index are provided, so that the multi-stage comprehensive toughness of the cross-domain unmanned cluster is comprehensively and systematically evaluated. According to the method, the defects in cross-domain unmanned cluster toughness evaluation in the prior art are overcome; through longitudinal analysis of a multilayer coupling failure mechanism and transverse construction of a multi-stage toughness process model, the performance change of the cross-domain unmanned cluster in different time stages can be more accurately grasped, so that powerful support is provided for improving the stability of the cluster and the task execution success rate.

Description

Technical Field

[0001] The present application relates to the field of resilience assessment technology, and in particular to a cross-domain unmanned cluster multi-stage comprehensive resilience assessment method, device and computer equipment. Background Art

[0002] With the development of unmanned platform technology in recent years, the concept of cross-domain collaborative operations has been proposed. The application of unmanned clusters in cross-domain collaborative operations has become more and more extensive. Unmanned platforms can achieve diversified tasks such as reconnaissance, interference, guidance, attack and service support through clustered operation. Its large-scale application and collaborative operation mode have become important development directions. This trend has put forward higher requirements on the resilience of unmanned clusters, especially in multi-stage operation tasks. Unmanned clusters need to have multi-stage resilience capabilities such as prevention, degradation, recovery and reconstruction to better cope with the situation of operation scenarios.

[0003] However, existing technologies have many deficiencies in the resilience assessment of cross-domain unmanned clusters, and cannot comprehensively and accurately assess the response capabilities of cross-domain unmanned clusters at different stages of operations. Current research on unmanned clusters mainly focuses on single-domain clusters or homogeneous clusters, while there is less discussion on cross-domain collaboration. With the development of unmanned platform technology, the complexity and heterogeneity of operating nodes in unmanned clusters and the diversity of relationships between nodes are becoming increasingly significant. In addition, most existing studies on resilience assessment characterize resilience from the overall performance changes of single-domain or homogeneous unmanned clusters or unmanned systems in the resilience process, and rarely measure the resilience indicators corresponding to the subdivided stages of different resilience processes of cross-domain unmanned clusters. Comprehensively modeling the resilience of cross-domain unmanned clusters in multiple time dimensions can quantitatively assess the prevention, resistance, recovery and reconstruction capabilities of clusters, ensuring that clusters can effectively respond to complex situations such as environmental changes and task changes at different stages and maintain efficient and stable operation.

[0004] Reference 1 (C. Zhang, T. Liu, G. Bai, J. Tao, and W. Zhu, “A dynamic resilience evaluation method for cross - domain swarms in confrontation,” Reliab. Eng. Syst. Saf., vol. 244, p. 109904, Apr. 2024.) proposed a dynamic resilience evaluation method for cross - domain swarms in a confrontation environment. Although this paper has made valuable contributions to the dynamic resilience evaluation of cross - domain unmanned swarms, there are still deficiencies in the analysis of multi - layer coupling failure mechanisms and multi - stage resilience evaluation. This paper mainly focuses on the survival rate and resilience evaluation of swarms in a confrontation environment, but lacks in - depth analysis of the multi - layer coupling failure mechanisms within cross - domain swarms. For example, cross - domain swarms involve multiple spatial domains, and there may be complex coupling relationships in communication, collaboration, and information transfer between different domains. These relationships may lead to chain reactions during failure, but the paper does not explore them in depth. In addition, the dynamic resilience evaluation method proposed in this paper mainly focuses on the real - time performance and recovery ability of swarms during confrontation, but lacks systematic research on the resilience evaluation of swarms in different stages (such as prevention, degradation, recovery, reconstruction stages, etc.). The survival rate may be a key indicator in a confrontation environment, but it may not be the most appropriate indicator in the prevention stage or non - confrontation environment. The evaluation of dynamic resilience in this paper describes resilience from the overall performance changes in the system resilience process, without dividing the specific time stages of the resilience process, and without separately evaluating the resilience indicators from different stages of resilience, making it impossible to measure the different resilience performances of swarms in various resilience stages in a changing environment in detail.

[0005] In summary, the shortcomings of the existing evaluation techniques and the reasons for their existence are analyzed as follows: (1) Currently, there is no method that simultaneously considers the spatial elements of cross - domain collaboration longitudinally and the time elements of the resilience process horizontally to evaluate the multi - stage resilience of cross - domain unmanned swarms. Cross - domain unmanned swarms face multiple threats such as communication interference, unmanned platform degradation, and cyber attacks in a complex environment. If a single event in one spatial domain triggers a failure, such as a broken communication link or insufficient fuel for an unmanned platform, it can cause faults to spread across different domains through the inter - layer interconnected communication links, resulting in cascading failures. The resilience of cross - domain unmanned swarms is affected not only by the connectivity between unmanned platform nodes within each spatial domain and between unmanned platform nodes in different spatial domains, but also by the time characteristics during task execution. Existing technologies lack an evaluation method that simultaneously considers spatio - temporal elements and cannot comprehensively evaluate the multi - stage comprehensive resilience of cross - domain unmanned swarms.

[0006] (2) Most of the existing technologies do not divide the resilience intermediate process into stages, and it is impossible to measure the different resilience performances of each resilience stage of the cross-domain unmanned cluster in a changing environment in detail. The resilience process involves multiple stages, and the performance changes and requirements of each stage are different. The existing technologies lack detailed division and measurement methods for these stages and cannot comprehensively evaluate the resilience performance of the cross-domain unmanned cluster. Summary of the Invention

[0007] Based on this, in view of the above technical problems, it is necessary to provide a multi-stage comprehensive resilience evaluation method, device and computer equipment for cross-domain unmanned clusters.

[0008] A multi-stage comprehensive resilience evaluation method for cross-domain unmanned clusters, the method includes: According to the given cross-domain unmanned cluster, construct a multi-layer heterogeneous network of the cross-domain unmanned cluster, and longitudinally analyze the multi-layer coupling failure mechanism of the cross-domain unmanned cluster according to the connectivity between unmanned platform nodes in different spatial domains, and determine the resilience strategy for resilience process modeling.

[0009] Horizontally consider multiple time periods during the task execution process, divide the resilience process of the cross-domain unmanned cluster into a prevention stage, a degradation stage, a recovery stage and a reconstruction stage, construct a resilience process model of the cross-domain unmanned cluster, and respectively construct the resilience indicators corresponding to each stage.

[0010] According to the resilience indicators of the prevention stage, the degradation stage, the recovery stage and the reconstruction stage, construct a multi-stage comprehensive resilience model of the cross-domain unmanned cluster.

[0011] Adopt the multi-stage comprehensive resilience model of the cross-domain unmanned cluster, and use the number of kill chains as the performance index to evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster to be evaluated in different scenarios.

[0012] A multi-stage comprehensive resilience evaluation device for cross-domain unmanned clusters, the device includes: A multi-layer coupling failure mechanism analysis module, configured to construct a multi-layer heterogeneous network of the cross-domain unmanned cluster according to the given cross-domain unmanned cluster, and longitudinally analyze the multi-layer coupling failure mechanism of the cross-domain unmanned cluster according to the connectivity between unmanned platform nodes in different spatial domains, and determine the resilience strategy for resilience process modeling.

[0013] A resilience indicator determination module, configured to horizontally consider multiple time periods during the task execution process, divide the resilience process of the cross-domain unmanned cluster into a prevention stage, a degradation stage, a recovery stage and a reconstruction stage, construct a resilience process model of the cross-domain unmanned cluster, and respectively construct the resilience indicators corresponding to each stage.

[0014] The cross - domain unmanned cluster multi - stage comprehensive resilience model construction module is used to construct a multi - stage comprehensive resilience model of the cross - domain unmanned cluster according to the resilience indicators in the prevention stage, degradation stage, recovery stage, and reconstruction stage.

[0015] The cross - domain unmanned cluster multi - stage comprehensive resilience evaluation module is used to adopt the multi - stage comprehensive resilience model of the cross - domain unmanned cluster and take the number of kill chains as the performance index to evaluate the multi - stage comprehensive resilience of the cross - domain unmanned cluster to be evaluated in different scenarios.

[0016] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of any of the above - mentioned methods are implemented.

[0017] For the above - mentioned cross - domain unmanned cluster multi - stage comprehensive resilience evaluation method, device, and computer device, the method deeply analyzes the multi - layer coupling failure mechanism of the cross - domain unmanned cluster, constructs a resilience process model covering four stages of prevention, degradation, recovery, and reconstruction, and proposes specific measurement methods for corresponding preventive indicators, robustness indicators, recoverability indicators, and reconfigurability indicators, so as to comprehensively and systematically evaluate the multi - stage comprehensive resilience of the cross - domain unmanned cluster. This method makes up for the deficiencies in the resilience evaluation of cross - domain unmanned clusters in the existing technology; by longitudinally analyzing the multi - layer coupling failure mechanism and horizontally constructing a multi - stage resilience process model, it can more accurately grasp the performance changes of the cross - domain unmanned cluster at different time stages, thus providing strong support for improving the stability of the cluster and the success rate of task execution. Description of the Drawings

[0018] Figure 1 It is a schematic flow chart of the cross - domain unmanned cluster multi - stage comprehensive resilience evaluation method in an embodiment; Figure 2 It is a schematic diagram of multi - layer coupling failure propagation of the cross - domain intelligent unmanned cluster in another embodiment Figure 3 It is a schematic diagram of the resilience strategy of the cross - domain intelligent unmanned cluster in another embodiment; Figure 4 It is a schematic diagram of the multi - stage resilience evolution process of the cross - domain unmanned cluster in another embodiment; Figure 5 It is a schematic diagram of the positions of each unmanned platform node in the space domain in another embodiment; Figure 6 It is a curve graph of the change in the number of kill chains of the cross - domain unmanned cluster without considering recovery and reconstruction strategies in another embodiment; Figure 7 It is a resilience curve graph of the cross - domain unmanned cluster in an adversarial environment in another embodiment; Figure 8Schematic diagram of the multi-stage comprehensive resilience of a cross-domain unmanned cluster under the influence of different reconstruction strategies in an adversarial environment for another embodiment; Figure 9 Resilience curve of a cross-domain unmanned cluster in a non-adversarial environment for another embodiment; Figure 10 Schematic diagram of the multi-stage comprehensive resilience of a cross-domain unmanned cluster under the influence of different reconstruction strategies in a non-adversarial environment for another embodiment; Figure 11 Internal structure diagram of a computer device in one embodiment. Detailed implementation manners

[0019] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0020] The multi-stage comprehensive resilience evaluation method for a cross-domain unmanned cluster proposed by the present application first, according to the given cross-domain unmanned cluster, considering the connectivity between unmanned platform nodes in different spatial domains, longitudinally analyzes the multi-layer coupling failure mechanism of the cross-domain unmanned cluster, and at the same time gives possible resilience strategies to prepare for the resilience process modeling. Secondly, considering multiple time stages during the task execution process horizontally, constructs a resilience process model of the cross-domain unmanned cluster, divides the resilience process of the cross-domain unmanned cluster into four stages: prevention, degradation, recovery, and reconstruction, and respectively proposes measurement methods for the corresponding preventive, robust, recoverable, and reconstructable resilience indicators for the four stages. Thirdly, constructs a resilience indicator system for the cross-domain unmanned cluster to evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster. Finally, uses case simulation research to illustrate the accuracy and effectiveness of the proposed method. In addition, the resilience indicator measurement methods and resilience indicator system proposed for each stage can provide a scientific basis for the optimal design and performance improvement of the cluster system, and further promote the development and application of cross-domain unmanned cluster technology.

[0021] In one embodiment, as Figure 1 shown, a multi-stage comprehensive resilience evaluation method for a cross-domain unmanned cluster is provided, and the method includes the following steps: Step 100: According to the given cross-domain unmanned cluster, construct a multi-layer heterogeneous network of the cross-domain unmanned cluster, and longitudinally analyze the multi-layer coupling failure mechanism of the cross-domain unmanned cluster according to the connectivity between unmanned platform nodes in different spatial domains, and determine the resilience strategy for resilience process modeling.

[0022] Specifically, this method takes cross-domain unmanned clusters as the research object, which cover unmanned platforms on the ground, in the air, on the sea surface, and under the sea floor, etc. However, cross-domain unmanned clusters face multiple threats such as communication interference, node failures, and cyberattacks in complex and harsh environments. If a single event in one domain (such as a broken communication link or insufficient fuel of an unmanned platform) triggers a failure, it may cause cross-domain propagation of faults through the communication links interconnected between domains, thus triggering cascading failures. Therefore, it is necessary to perform resilient recovery and reconstruction on failed unmanned platforms. In a complex environment, the resilience of the cluster is directly related to the success or failure of the mission and affects the overall operation efficiency of the cross-domain unmanned cluster. Therefore, performing resilience modeling and evaluation on cross-domain unmanned clusters is of crucial significance for ensuring their stable and efficient operation.

[0023] Spatio-temporal characteristics refer to temporal characteristics and spatial features. In terms of time, it mainly means that the division of multi-stage resilience is carried out in chronological order, which is horizontal. From left to right in chronological order are prevention, degradation, recovery, and reconstruction. In terms of space, it mainly refers to the characteristics of the spatial distribution of cross-domain unmanned clusters, which cover four spatial domains and are vertical. From top to bottom are air, land, sea surface, and sea floor.

[0024] Driven by the same mission objective, cross-domain unmanned clusters operate in different spatial domains, and each unmanned platform cooperates with each other to achieve complementary functions through task planning, information interaction, behavior coordination, etc., so as to achieve mission success. According to the spatial domain division, cross-domain intelligent unmanned clusters can be modeled as an air unmanned platform system, a land unmanned platform system, a sea surface unmanned platform system, and a sea floor unmanned platform system. At the same time, cross-domain intelligent unmanned clusters can be modeled as an operation system, a communication system, a command system, and a support system according to their functions. The operation system includes reconnaissance and detection unmanned platforms and strike unmanned platforms. The operation system and the command system can complete information sharing through the communication system. The reconnaissance and detection unmanned platforms, command and decision-making unmanned platforms, and strike unmanned platforms and the target can form a kill chain, and the unmanned platforms in the communication system are mainly presented in the form of links between unmanned platforms in the kill chain. The support unmanned platforms in the support system are deployed in the operation system, communication system, and command system, which play an important role in improving the reliability of cross-domain intelligent unmanned clusters, realizing information sharing and real-time monitoring within the cluster, optimizing command and decision-making, and enhancing the overall efficiency. In an uncertain environment, various unmanned platforms may malfunction at any time, and the support system can provide maintenance services and resource replenishment for the operation system, communication system, and command system, injecting new vitality into the degraded cross-domain intelligent unmanned cluster.

[0025] In an adversarial scenario, unmanned platforms within the operating system are most vulnerable to being affected by the other party. The communication system is easily interfered with by the other party's electronic devices, resulting in chaotic signal transmission. As the core of the cluster, the command system is at great risk of being deliberately destroyed. In a non-adversarial scenario, various unmanned platforms may experience natural random failures due to internal reasons such as loose components, wear, and aging. Therefore, the failure modes of cross-domain unmanned clusters are mainly divided into two types: deliberate sabotage mode and random failure mode. Deliberate sabotage strategies include maximum-degree sabotage, random sabotage, etc. Random failure refers to the natural failures of unmanned platforms caused by equipment aging, loose components, etc. The material reserves of the safeguard system are of the cold backup nature, in an inactive state, not bearing the workload, and with extremely low wear and usage levels. Since the standby equipment is in an idle state, its lifespan will be relatively extended. The cross-domain intelligent unmanned cluster conducts multi-layer coupled failure propagation as Figure 2 shown.

[0026] Cascading failure refers to the phenomenon in a complex system where the failure of a certain component or node triggers the successive failures of other components or nodes, which may ultimately lead to the paralysis of the entire system or most of its functions. It mainly involves the failure propagation mechanism, that is: the initial failure propagates through the connection structure and interdependence of the system, resulting in an increase in the load, a decrease in performance, or a direct failure of other components or nodes.

[0027] When an unmanned platform fails, it can no longer complete its normal work and cannot transmit information to its front-end and back-end unmanned platform nodes. After an unmanned platform fails, the resilience strategy of the cross-domain unmanned cluster is as Figure 3 shown.

[0028] In the existing technology, the research on unmanned clusters often focuses on the collaboration and control within a single spatial domain. For cross-domain unmanned clusters, especially those involving the collaboration of multiple spatial domains, the connectivity between unmanned platform nodes in different spatial domains and their cascading failures are not considered comprehensively and deeply enough. During the actual execution of cross-domain tasks, when the communication between unmanned platform nodes in different spatial domains is restricted or connection failures occur, it is impossible to accurately estimate and effectively deal with the resulting decline in cluster performance, thus affecting the execution effect and success rate of the entire cluster task. This method constructs a multi-layer heterogeneous network of cross-domain unmanned clusters, fully considering the connectivity between unmanned platform nodes in different spatial domains, solves the deficiency of the existing technology in the failure mechanism analysis considering spatial characteristics, enables the unmanned cluster to better detect potential interferences during the execution of cross-domain tasks, and improves the stability and task execution success rate of the cluster in a complex spatial environment.

[0029] Step 102: Horizontally consider multiple time periods during the task execution process, divide the resilience process of the cross-domain unmanned cluster into a prevention stage, a degradation stage, a recovery stage, and a reconstruction stage, build a resilience process model of the cross-domain unmanned cluster, and build resilience indicators corresponding to each stage respectively.

[0030] Specifically, although some scattered resilience indicators have been proposed in the current research on the resilience of unmanned clusters, a complete indicator system specifically for the multi-stage comprehensive resilience of cross-domain unmanned clusters has not yet been formed. At the same time, there are also deficiencies in the measurement methods of resilience indicators, and there is a lack of effective measurement methods for resilience indicators such as preventability, robustness, recoverability, and reconfigurability corresponding to the four stages of prevention, degradation, recovery, and reconstruction of cross-domain unmanned clusters. This leads to the inability to accurately quantify and compare the resilience levels of different unmanned cluster systems or the same cluster system in different states in practical applications, which is not conducive to the optimal design and performance improvement of cluster systems.

[0031] This method constructs a cross-domain unmanned cluster resilience index system, improves the deficiencies of existing resilience indicators, and proposes an effective measurement method for each resilience indicator to solve the problems of existing technologies in resilience index system and measurement methods. By accurately quantifying and comparing the resilience level of cross-domain unmanned clusters under different failure modes or resilience strategies, it provides a scientific basis for the optimal design, performance evaluation and improvement of cluster systems, and promotes the development and application of cross-domain unmanned cluster technology.

[0032] The resilience process of cross-domain unmanned clusters is divided into four stages: prevention, degradation, recovery, and reconstruction. Figure 4 As shown in the figure, four resilience indicators are given for the performance of the four stages of the cross-domain unmanned cluster: preventability indicator, robustness indicator, recoverability indicator, and reconfigurability indicator. Figure 4 Medium performance It is expressed by the number of kill chains of cross-domain unmanned clusters.

[0033] Resilience refers to the ability of a system to maintain or quickly restore its functions and performance when it is disturbed or fails. The resilience of a cross-domain unmanned cluster refers to the comprehensive characteristics of an unmanned cluster being able to maintain or quickly restore its functions, performance, and mission execution capabilities in a complex and changing cross-domain environment when faced with external interference, failures, or changes in mission requirements. The method specifically includes the prevention, resistance, recovery, and reconstruction capabilities of a cross-domain unmanned cluster when faced with internal and external potential threats.

[0034] A resilience strategy refers to a strategy in a complex system that, in response to various disturbances, shocks, and uncertainties, enhances the resilience of the system (i.e., the ability to recover and continue operating in the face of adversity) by designing and implementing a series of measures. These strategies aim to improve the system's recovery ability and reconstruction ability when facing various challenges, ensuring that the system can quickly return to the normal state after being impacted and even achieve better performance. In the present invention, there are three resilience strategies. The first is the supply and repair strategy, which is used in the recovery stage of the multi-stage resilience process. The second and third are the intra-cluster and inter-cluster reconstruction strategies, which are used in the reconstruction stage of the multi-stage resilience process.

[0035] Step 104: Construct a multi-stage comprehensive resilience model for cross-domain unmanned clusters based on the resilience indicators in the prevention stage, degradation stage, recovery stage, and reconstruction stage.

[0036] Specifically, most of the existing technologies focus on the performance of unmanned clusters at a specific moment or stage. For multiple time stages during the task execution process, such as the prevention stage (anticipating and preventing potential risks), degradation stage (the process of performance decline when suffering from interference or failure), recovery stage (the process of recovering from the degraded state to a higher performance state), and reconstruction stage (re-adjusting and optimizing the cluster structure), there is a lack of constructing a systematic resilience process model. This makes it impossible to comprehensively and accurately evaluate the resilience performance of unmanned clusters at different stages and conduct targeted optimization in the face of complex and changeable task environments and various interference factors, and it is difficult to effectively manage and improve the cluster resilience throughout the task cycle. Therefore, constructing a resilience process model for cross-domain unmanned clusters, dividing the task execution process into four stages: prevention, degradation, recovery, and reconstruction, and proposing corresponding resilience indicator measurement methods for each stage can make up for the deficiencies of the existing technologies in multi-time-stage resilience assessment. This helps to comprehensively and accurately evaluate and monitor the resilience performance of unmanned clusters at each stage of task execution, timely discover problems and take targeted optimization measures, so as to effectively manage and improve the cluster resilience throughout the task cycle and ensure the smooth completion of the task.

[0037] Step 106: Use the multi-stage comprehensive resilience model of cross-domain unmanned clusters and take the number of kill chains as the performance indicator to evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster to be evaluated in different scenarios.

[0038] In the above cross-domain unmanned cluster multi-stage comprehensive resilience evaluation method, the method comprehensively and systematically evaluates the multi-stage comprehensive resilience of the cross-domain unmanned cluster by deeply analyzing the multi-layer coupling failure mechanism of the cross-domain unmanned cluster, constructing a resilience process model covering four stages of prevention, degradation, recovery, and reconstruction, and proposing specific measurement methods for the corresponding preventability index, robustness index, recoverability index, and reconfigurability index. This method makes up for the deficiencies of the existing technology in the resilience evaluation of cross-domain unmanned clusters; by vertically analyzing the multi-layer coupling failure mechanism and horizontally constructing a multi-stage resilience process model, it can more accurately grasp the performance changes of the cross-domain unmanned cluster at different time stages, thus providing strong support for improving the stability of the cluster and the success rate of task execution.

[0039] In one embodiment, the resilience strategies in step 100 include: a supply and repair strategy, an intra-cluster reconstruction strategy, and an inter-cluster reconstruction strategy; the supply and repair strategy is used in the recovery stage of the multi-stage resilience process, and the intra-cluster reconstruction strategy and the inter-cluster reconstruction strategy are used in the reconstruction stage of the multi-stage resilience process.

[0040] The supply and repair strategy is used to ensure that when a certain unmanned platform has insufficient supplies, the system can use internal resources to replenish the resources of the unmanned platform. If the unmanned platform suffers a minor failure, the system can repair the unmanned platform to ensure the stable and safe operation of the cross-domain intelligent unmanned cluster.

[0041] The intra-cluster reconstruction strategy is used when a certain unmanned platform in the cross-domain unmanned cluster is deliberately damaged and fails or suffers a natural random failure. Similar unmanned platforms in the same cluster as the failed unmanned platform can replace the failed unmanned platform to work; among them, the connection of the communication link between the new unmanned platform and the unmanned platform connected to the failed node is involved in the substitution work process.

[0042] The inter-cluster reconstruction strategy is used when a certain unmanned platform in the cross-domain unmanned cluster is deliberately damaged and fails or suffers a natural random failure. Similar unmanned platforms in other unmanned clusters can replace the failed unmanned platform to execute tasks; among them, the connection of the communication link between the new unmanned platform and the unmanned platform connected to the failed node is involved in the substitution work process.

[0043] In one embodiment, the expression of the preventability index corresponding to the prevention stage in step 102 is:

[0044] Wherein, is the preventability index of the cross-domain unmanned cluster, is the risk prediction ability, is the early warning ability, is the pre-event response ability.

[0045] Specifically, the prevention stage refers to the stage before the cross-domain unmanned cluster is interfered by risks. In this stage, the performance of the cross-domain unmanned cluster remains unchanged. Once a risk event that hinders the task completion is detected, the cross-domain unmanned cluster will initiate the corresponding pre-plan in advance and take certain actions for pre-resistance. If the impact of the risk event is completely suppressed, the performance of the cross-domain unmanned cluster remains unchanged; otherwise, the performance begins to decline and the cross-domain unmanned cluster enters the degradation stage.

[0046] In the prevention stage, preventive indicators are used to evaluate this resilience process, indicating the ability of the cross-domain unmanned cluster to autonomously prevent risks and prevent their adverse effects on the cluster in advance. The prevention ability of the cross-domain unmanned cluster needs to be based on accurate prediction and judgment analysis of deliberate sabotage or internal degradation in order to take effective protection measures in a timely manner. If the cross-domain unmanned cluster can accurately predict the occurrence and possible impact range of risk events and strengthen the monitoring, early warning, and response capabilities of the protected area, the losses caused by risks can be reduced. Similarly, if the cross-domain unmanned cluster can regularly take measures for preventive maintenance of unmanned platforms, the uncertain impacts brought by the natural aging of unmanned platforms can also be avoided. Based on this, the evaluation of the preventive indicators of the cross-domain unmanned cluster needs to consider the risk prediction ability , early warning ability and pre-response ability . Therefore, the preventive indicators of the cross-domain unmanned cluster are as shown in the expression of the above preventive indicators.

[0047] In one embodiment, the expression of the robustness indicator corresponding to the degradation stage in step 102 is:

[0048] Wherein, is the robustness indicator of the cross-domain unmanned cluster, is the resistance coefficient, and the resistance coefficient is the ratio of the number of normal devices in the cross-domain unmanned cluster before and after resisting interference in the degradation stage; is the redundancy coefficient based on the task baseline, and the redundancy coefficient is the ratio of the part where the performance of the cross-domain unmanned cluster exceeds the task baseline to the task baseline, is the traditional resilience in the degradation stage.

[0049] Specifically, after deliberate sabotage or random failure occurs and the cross-domain unmanned cluster cannot prevent it, the cross-domain unmanned cluster will quickly enter the resistance state. Although it cannot completely resist the impact of interference, it can weaken the impact of risks on the cross-domain unmanned cluster to a certain extent. Under the combined action of interference and the cluster's resistance behavior, the performance of the cross-domain unmanned cluster will drop to a certain threshold. As Figure 3 shown, in At a moment, if the impact intensity of a severe event is greater than the prevention ability of the cross - domain unmanned cluster, the performance of the unmanned cluster degrades. During the period from to , the resistance ability of the cluster can offset part of the impact brought by the risk. At the moment of , the cluster supply and repair strategy is formulated, and sufficient preparations have been made for the cross - domain unmanned cluster to enter the recovery stage. At this moment, the cluster performance drops to the threshold and then starts to rise.

[0050] The typical calculation method of traditional resilience is: the ratio of the area enclosed by the performance curve and the horizontal axis during a certain period under interference to the area enclosed by the performance and the horizontal axis without interference. Therefore, the traditional resilience in the degradation stage is , where is the function of performance changing with time, and is the initial performance of the cross - domain unmanned cluster. Considering the robustness and redundancy requirements of the cross - domain unmanned cluster, a resistance coefficient and a redundancy coefficient are introduced to represent the ability of the cluster to still execute tasks during the degradation stage. The resistance coefficient is defined as the ratio of the number of normal devices in the cross - domain unmanned cluster before and after resisting interference during the degradation stage, and is represented by . Redundancy describes the replaceable functions and designs in the operation of the cross - domain unmanned cluster. Redundancy design is to better complete tasks, which enables the cross - domain unmanned cluster to maintain a performance level beyond the task baseline. Therefore, a redundancy coefficient based on the task baseline is introduced, which is defined as the ratio of the part where the performance of the cross - domain unmanned cluster exceeds the task baseline to the task baseline. The larger , the higher the redundancy of the cross - domain unmanned cluster, which has a certain positive effect on the resistance ability of the cluster. Therefore, the robustness index

[0051] of the cross - domain unmanned cluster is as shown in the above - mentioned expression of the robustness index.

[0052]

[0053] In one of the embodiments, the recoverability index corresponding to the recovery stage in step 102 is: , where is the recovery coefficient, is the value of the traditional resilience in the recovery stage, is the value of the function of performance changing with time at the moment of , and

[0054] Figure 4 Specifically, in the recovery stage, if the recovery intensity of the cross - domain unmanned cluster is greater than the risk interference intensity, the performance of the cross - domain unmanned cluster will continue to improve based on the minimum value. As shown in Figure 4 At this moment, the cross-domain unmanned cluster starts to implement the resilience strategy 1, that is, the supply and repair strategy, and the cluster performance continuously improves rapidly until At this moment, the recovery of the unmanned cluster performance is close to the saturation state, and the performance improvement speed suddenly decreases. At this time, even if similar supply and repair measures are continued, the cluster performance still rises slowly.

[0055] In the recovery stage, a recoverability index is used to evaluate this resilience process, indicating the degree or speed of the recoverability of the cross-domain unmanned cluster performance after degradation, and is represented by . The traditional resilience value in the recovery stage is , and the measurement method is similar to the traditional resilience in the degradation stage, and the method of area ratio is also used. In the actual operation process of the cross-domain unmanned cluster, the end point of the recovery time is often uncertain, and the focus of this method is whether the cross-domain unmanned cluster can complete the mission successfully. Therefore, considering the influence of the task baseline on the mission completion of the cross-domain unmanned cluster, a recovery coefficient is introduced, The larger the value, the stronger the recovery ability of the cross-domain unmanned cluster, which has a positive effect on the improvement of the cluster resilience.

[0056] In one of the embodiments, the expression of the reconfigurability index corresponding to the reconstruction stage in step 102 is:

[0057]

[0058] where is the reconfigurability index corresponding to the reconstruction stage, is the reconstruction coefficient, is the performance end point of the cross-domain unmanned cluster reconstruction, is the initial performance of the cross-domain unmanned cluster, is the value of the function of performance changing with time at moment, is the task baseline.

[0059] Specifically, the cross-domain unmanned cluster has not recovered to the expected performance at moment, and thus major adjustments to the cluster structure or composition are required, and resilience strategy 2 or resilience strategy 3 is adopted, that is, in-cluster reconstruction and out-of-cluster reconstruction strategies, to promote the cross-domain unmanned cluster to break through the saturation state in the recovery stage and make its performance reach the original state or even higher. Figure 3 In, the cross-domain unmanned cluster ends the reconstruction at moment and reaches the performance value . So far, the cross-domain unmanned cluster completes the resilience process and starts to operate normally.

[0060] In the reconstruction stage, a reconfigurability index is used to evaluate the resilience process, indicating the structural reorganization ability of the cross-domain unmanned cluster to cope with risks, denoted by . The traditional resilience value in the reconstruction stage is , and the measurement method is similar to the traditional resilience in the degradation stage, also using the method of area ratio. In the actual operation process of the cross-domain unmanned cluster, reconstruction is not always required. Considering the differences in the performance starting point and ending point during reconstruction, is introduced to represent the reconstruction coefficient of the cross-domain unmanned cluster, . When in the recovery stage , if the cross-domain unmanned cluster can meet the task requirements without reconstruction, the reconstruction coefficient is set to . When , the reconstruction strategy of the cross-domain unmanned cluster needs to be considered. However, the performance ending point of the cross-domain unmanned cluster reconstruction is often uncertain. It may still be less than the task baseline , or the reconstruction effect may be excellent, making . Therefore, considering the impact of the task baseline and initial performance of the cross-domain unmanned cluster on resilience, the reconstruction coefficient is defined, and the reconfigurability index of the cross-domain unmanned cluster is as shown in the expression of the reconfigurability index.

[0061] In one embodiment, step 104 includes: constructing a multi-stage comprehensive resilience model of the cross-domain unmanned cluster according to the resilience indicators in the prevention stage, degradation stage, recovery stage, and reconstruction stage; the expression of the multi-stage comprehensive resilience model of the cross-domain unmanned cluster is:[[]]

[0062] wherein, is the multi-stage comprehensive resilience, and its value range is [0,1]; , , , are the preventability index, robustness index, recoverability index, and reconfigurability index of the cross-domain unmanned cluster respectively, and their value ranges are all [0,1].[[]]

[0063] Specifically, according to the four resilience processes of the prevention stage, degradation stage, recovery stage, and reconstruction stage, the multi-stage resilience index system of the cross-domain unmanned cluster includes the preventability index, robustness index, recoverability index, and reconfigurability index. Therefore, the multi-stage comprehensive resilience model of the cross-domain unmanned cluster is as shown in the expression of the above multi-stage comprehensive resilience model of the cross-domain unmanned cluster.[[]] The closer it is to 0, the weaker the abilities of the cross-domain unmanned cluster to prevent, resist, recover, and reconstruct when facing risks, the lower the cluster performance, and the worse the multi-stage comprehensive resilience.[[]] It indicates that when facing risk interference, the performance of the cross-domain unmanned cluster has no loss, and the multi-stage comprehensive resilience is the greatest.

[0064] In one embodiment, the failure modes of the cross-domain unmanned cluster include: deliberate sabotage failure, natural random failure, and hybrid failure; Deliberate sabotage failure includes maximum-degree sabotage and random sabotage.

[0065] Natural random failure refers to equipment aging, component wear and looseness.

[0066] Hybrid failure is a failure mode under the combined action of natural random failure and deliberate sabotage failure.

[0067] In one embodiment, the multi-stage resilience assessment process of the cross-domain unmanned cluster is as follows: Step 1: Given a cross-domain unmanned cluster, conduct a multi-layer coupled failure analysis on it.

[0068] Step 2: Construct a kill chain to form a kill network of the cross-domain unmanned cluster, calculate the number of kill chains of the cluster, and use the number of kill chains of the cluster as the performance of the cross-domain unmanned cluster.

[0069] Step 3: Consider deliberate sabotage failure strategies, natural random failure strategies, supply and repair strategies, in-cluster reconstruction strategies, and inter-cluster reconstruction strategies, etc., and analyze the multi-stage resilience process of the cross-domain unmanned cluster.

[0070] Step 4: Quantitatively evaluate the multi-stage resilience indicators of the cross-domain unmanned cluster - preventability indicator, robustness indicator, recoverability indicator, and reconfigurability indicator.

[0071] Step 5: Quantitatively evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster.

[0072] It should be understood that although Figure 1 the steps in the flowchart of Figure 1 are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover,

[0073] In a verification example, to verify the applicability of the method, taking the case of an unmanned cluster performing tasks in different spatial domains as an example, and using the number of kill chains as a performance indicator, the multi-stage resilience of the cross-domain unmanned cluster in different scenarios is evaluated. There are 6 types of unmanned clusters performing tasks. Each unmanned cluster can work independently and can also cooperate with other clusters to jointly perform complex tasks. Each cluster has 1 set of command systems, and each set of command systems supports the access of 4 detection unmanned platforms and 8 strike unmanned platforms. The 6 unmanned clusters in total have 24 reconnaissance unmanned platforms, 6 decision-making unmanned platforms, and 48 strike unmanned platforms. Assuming that the unmanned platforms in all unmanned clusters cooperate to confront 12 targets of the other party, and the number distribution of the targets in the four spatial domains is 3 each. Furthermore, the unmanned platform system composed of these 6 unmanned clusters can be abstracted as a multi-layer heterogeneous directed network composed of 90 nodes, and the positions of the unmanned platform nodes in the spatial domain are as Figure 5 shown.

[0074] To make the simulation more in line with the actual task scenario, the task baseline is set to 0.8 times the initial performance. At the same time, considering the ability differences of unmanned platforms in different spatial domains, ability attributes are added to each of our unmanned platforms. The ability parameters of unmanned platforms and communication links in different spatial domains are shown in Table 1.

[0075] Table 1 Ability parameters of unmanned platforms and communication links in different spatial domains

[0076] (1)Failure analysis of cross-domain unmanned clusters During the multi-stage resilience process, in the degradation stage, natural random failures, deliberate sabotage failures by the other party, and degradations of unmanned platforms due to insufficient supplies may occur within the cross-domain unmanned cluster. Among them, the natural random failure strategy refers to the failure of unmanned platforms caused by internal reasons, and the deliberate sabotage failure strategy takes into account the sabotage methods of the other party during the confrontation between the two sides. From the perspective of the other party, it will preferentially target the unmanned platform with the largest node degree on our side for deliberate sabotage. The mixed failure strategy is the failure mode under the combined action of the natural random failure strategy and the deliberate sabotage failure strategy. In the recovery stage, the cross-domain unmanned cluster adopts a supply and repair strategy for the degraded unmanned platforms, that is, the basic capabilities of the cluster are restored through the timely supply of supplies. In the reconstruction stage, the cross-domain unmanned cluster adopts strategies of intra-cluster reconstruction and inter-cluster reconstruction for the failed unmanned platforms to improve the task execution ability of the cluster. In the confrontation environment, a mixed failure strategy considering the combined action of the natural random failure strategy and the deliberate sabotage failure strategy is considered; in the non-confrontation environment, only the natural random failure strategy is considered.

[0077] When not considering the recovery and reconstruction strategies, the influence of different failure strategies on the number of kill chains of the cross-domain unmanned cluster is as Figure 6 shown.

[0078] Figure 6 Among them, the total number of cross-domain unmanned cluster kill chains at the initial moment is 3,280. Under the action of different failure strategies, the number of kill chains shows an obvious downward trend. Under the action of the natural random failure strategy, the decline rate of the number of cross-domain unmanned cluster kill chains changes from slow to fast. At the 100th hour, the number of kill chains drops to 261. Under the action of the deliberate sabotage failure strategy, the number of cross-domain unmanned cluster kill chains drops relatively fast. At the 100th hour, the number of kill chains drops to 85. Under the action of the hybrid failure strategy, the number of cross-domain unmanned cluster kill chains drops the fastest. At the 100th hour, the number of kill chains drops to 2. Therefore, it can be seen that the hybrid failure strategy has the greatest impact on the number of cross-domain unmanned cluster kill chains, followed by the deliberate sabotage failure strategy, and the natural random failure strategy has the smallest impact. This is because the natural random failure strategy is caused by internal degradation, component loosening, wear, etc. of the unmanned platform. The natural random failure rate of the unmanned platform is small, so the natural random failure of the unmanned platform has a small impact on the number of cross-domain unmanned cluster kill chains. The deliberate sabotage failure strategy is deliberately carried out by the other party, with strong aggressiveness. It preferentially attacks the nodes with the greatest degree of our unmanned platforms, and has a greater lethality. Therefore, the deliberate sabotage failure of the unmanned platform has a greater impact on the number of cross-domain unmanned cluster kill chains. The hybrid failure strategy is the combined action of the two natural random failure strategies and the deliberate sabotage failure strategy, and its impact on the number of cross-domain unmanned cluster kill chains will be greater than that of a single failure strategy.

[0079] (2)Analysis of the resilience of cross-domain unmanned clusters under the influence of different reconstruction strategies in the confrontation environment When considering recovery and reconstruction strategies, since the carrying capacity of unmanned platforms is limited and the workload is not infinite, therefore, the cross-domain unmanned cluster will preferentially select the unmanned platform with the smallest degree among the non-failed unmanned platforms for reconstruction. Considering the limitations of reconstruction resources and cluster control capabilities, it is assumed that the cross-domain unmanned cluster can only reconstruct 3 failed unmanned platforms at a time. The resilience curves under the in-cluster reconstruction strategy, the inter-cluster reconstruction strategy, and the hybrid reconstruction strategy (both in-cluster and inter-cluster can be reconstructed) in the confrontation environment (i.e., under the action of the hybrid failure strategy) are as Figure 7 shown.

[0080] Figure 7Among them, from 0 to 10 hours, the cross-domain unmanned cluster is in the prevention stage, the number of kill chains remains at the initial value, and the cluster maintains its initial performance state. From 10 to 40 hours, the cross-domain unmanned cluster is affected by the hybrid failure strategy, and the number of kill chains gradually decreases. Due to the resistance of the cross-domain unmanned cluster, the number of kill chains does not drop to 0. During this period, the cross-domain unmanned cluster is in a degraded state. From 40 to 70 hours, the cross-domain unmanned cluster begins to implement supply and repair measures, and the number of kill chains gradually recovers. During this period, the cross-domain unmanned cluster is in the recovery stage. From 70 to 80 hours, the cross-domain unmanned cluster executes the reconstruction strategy. During the reconstruction stage, the growth rate of the number of kill chains is faster than that in the recovery stage, enabling the cluster to increase the number of kill chains to the maximum value under the current damaged state of some unmanned platforms within 10 hours.

[0081] The multi-stage comprehensive resilience of the cross-domain unmanned cluster under the influence of different reconstruction strategies in the adversarial environment is as Figure 8 shown.

[0082] As Figure 8 can be seen, considering the hybrid reconstruction strategy, the intra-cluster reconstruction strategy, and the inter-cluster reconstruction strategy, the resilience of the cross-domain unmanned cluster decreases in turn, indicating that the effects of the intra-cluster reconstruction strategy and the hybrid reconstruction strategy are better than the inter-cluster reconstruction effect in the adversarial environment, and the effect of the hybrid reconstruction strategy is slightly better than that of the intra-cluster reconstruction strategy.

[0083] (3)Analysis of the Resilience of the Cross-Domain Unmanned Cluster under the Influence of Different Reconstruction Strategies in the Non-Adversarial Environment In the non-adversarial environment, only natural random failures of our unmanned platforms occur during the degradation stage. In the non-adversarial environment, the resilience curves under the intra-cluster reconstruction strategy, the inter-cluster reconstruction strategy, and the hybrid reconstruction strategy (both intra-cluster and inter-cluster can be reconstructed) are as Figure 9 shown.

[0084] Figure 9 Among them, from 0 to 10 hours, the cross-domain unmanned cluster is in the prevention stage, and the number of kill chains remains at the initial value. From 10 to 40 hours is the degradation stage of the cross-domain unmanned cluster. Affected by the natural random failure strategy, the number of kill chains gradually decreases. The number of kill chains at the 40th moment is about 2,237. From 40 to 70 hours is the recovery stage of the cross-domain unmanned cluster. The cluster begins to implement supply and repair measures, and the number of kill chains gradually recovers. By the 70th hour, the number of kill chains is about 2,649. From 70 to 80 hours is the reconstruction stage of the cross-domain unmanned cluster. The cluster executes the reconstruction strategy to further increase the number of kill chains. The difference in the number of kill chains after the intra-cluster reconstruction strategy and the hybrid reconstruction strategy is not significant. The final number of kill chains under the hybrid reconstruction strategy is slightly higher than that under the intra-cluster reconstruction strategy, but both are significantly higher than the number of kill chains after the inter-cluster reconstruction strategy.

[0085] The multi-stage comprehensive resilience of cross-domain unmanned clusters under the influence of different reconstruction strategies in a non-confrontational environment is as follows Figure 10 as shown

[0086] According to Figure 10 it can be seen that the resilience of cross-domain unmanned clusters considering hybrid reconstruction strategies, intra-cluster reconstruction strategies, and inter-cluster reconstruction strategies decreases in turn, indicating that the effects of intra-cluster reconstruction strategies and hybrid reconstruction strategies in cross-domain unmanned clusters in a non-confrontational environment are better than those of inter-cluster reconstruction, and the effect of the hybrid reconstruction strategy is slightly better than that of the intra-cluster reconstruction strategy.

[0087] In one embodiment, a multi-stage comprehensive resilience evaluation device for cross-domain unmanned clusters is provided, including: a multi-layer coupled failure mechanism analysis module, a resilience index determination module, a multi-stage comprehensive resilience model construction module for cross-domain unmanned clusters, and a multi-stage comprehensive resilience evaluation module for cross-domain unmanned clusters, where The multi-layer coupled failure mechanism analysis module is used to construct a multi-layer heterogeneous network of the cross-domain unmanned cluster according to the given cross-domain unmanned cluster, and longitudinally analyze the multi-layer coupled failure mechanism of the cross-domain unmanned cluster according to the connectivity between unmanned platform nodes in different spatial domains, so as to determine the resilience strategy for resilience process modeling.

[0088] The resilience index determination module is used to horizontally consider multiple time periods during the task execution process, divide the resilience process of the cross-domain unmanned cluster into a prevention stage, a degradation stage, a recovery stage, and a reconstruction stage, construct a resilience process model of the cross-domain unmanned cluster, and respectively construct resilience indices corresponding to each stage.

[0089] The multi-stage comprehensive resilience model construction module for cross-domain unmanned clusters is used to construct a multi-stage comprehensive resilience model of the cross-domain unmanned cluster according to the resilience indices of the prevention stage, the degradation stage, the recovery stage, and the reconstruction stage.

[0090] The multi-stage comprehensive resilience evaluation module for cross-domain unmanned clusters is used to adopt the multi-stage comprehensive resilience model of the cross-domain unmanned cluster, and use the number of kill chains as a performance index to evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster to be evaluated in different scenarios.

[0091] In one of the embodiments, the resilience strategies in the multi-layer coupled failure mechanism analysis module include: replenishment and repair strategies, intra-cluster reconstruction strategies, and inter-cluster reconstruction strategies; the replenishment and repair strategies are used in the recovery stage of the multi-stage resilience process, and the intra-cluster reconstruction strategies and inter-cluster reconstruction strategies are used in the reconstruction stage of the multi-stage resilience process.

[0092] Supply repair strategy, which is used to ensure that the system can use internal resources to supply resources to a certain unmanned platform when the supplies of the unmanned platform are insufficient. If the unmanned platform suffers from minor faults, the guarantee system can repair the unmanned platform to ensure the stable and safe operation of the cross-domain intelligent unmanned cluster.

[0093] Intra-cluster reconstruction strategy, which is used when a certain unmanned platform in the cross-domain unmanned cluster is deliberately damaged and fails or undergoes natural random failure, and the same type of unmanned platform in the same cluster as the failed unmanned platform can replace the failed unmanned platform to work; among them, the connection of the communication link between the new unmanned platform and the unmanned platform connected to the failed node is involved in the substitution work process.

[0094] Inter-cluster reconstruction strategy, which is used when a certain unmanned platform in the cross-domain unmanned cluster is deliberately damaged and fails or undergoes natural random failure, and the same type of unmanned platform in other unmanned clusters can replace the failed unmanned platform to perform tasks; among them, the connection of the communication link between the new unmanned platform and the unmanned platform connected to the failed node is involved in the substitution work process.

[0095] In one embodiment, the preventability index corresponding to the prevention stage in the resilience index determination module is as shown in the above preventability index expression.

[0096] In one embodiment, the robustness index corresponding to the degradation stage in the resilience index determination module is as shown in the above robustness index expression.

[0097] In one embodiment, the recoverability index corresponding to the recovery stage in the resilience index determination module is as shown in the above recoverability index expression.

[0098] In one embodiment, the reconfigurability index corresponding to the reconstruction stage in the resilience index determination module is as shown in the above reconfigurability index expression.

[0099] In one embodiment, the cross-domain unmanned cluster multi-stage comprehensive resilience model construction module is further used to construct a cross-domain unmanned cluster multi-stage comprehensive resilience model as shown in the above cross-domain unmanned cluster multi-stage comprehensive resilience model expression according to the resilience indexes of the prevention stage, degradation stage, recovery stage, and reconstruction stage.

[0100] In one embodiment, the failure modes of the cross-domain unmanned cluster in the device include: deliberate damage failure, natural random failure, and mixed failure; deliberate damage failure includes maximum degree damage and random damage; natural random failure refers to equipment aging and component wear and looseness; mixed failure is a failure mode under the combined action of natural random failure and deliberate damage failure.

[0101] For the specific limitations of the cross - domain unmanned cluster multi - stage comprehensive resilience evaluation device, reference can be made to the limitations of the cross - domain unmanned cluster multi - stage comprehensive resilience evaluation method in the above text, which will not be elaborated here. Each module in the above - mentioned cross - domain unmanned cluster multi - stage comprehensive resilience evaluation device can be implemented in whole or in part by software, hardware, or a combination thereof. The above - mentioned modules can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so as to facilitate the processor to call and execute the operations corresponding to each of the above - mentioned modules.

[0102] In one embodiment, a computer device is provided. This computer device can be a terminal, and its internal structure diagram can be as Figure 11 shown. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus. 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 and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non - volatile storage medium. The network 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, it realizes a cross - domain unmanned cluster multi - stage comprehensive resilience evaluation method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0103] Those skilled in the art can understand that Figure 11 the structure shown in

[0104] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0104] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps in the above - mentioned method embodiment.

[0105] The technical features of the above - mentioned embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above - mentioned embodiments are described. However, as long as the combination of these technical features does not conflict, it should be considered to be within the scope described in this specification.

[0106] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A cross-domain unmanned cluster multi-stage comprehensive resilience assessment method, characterized in that The method includes: Construct a multi-layer heterogeneous network of the cross-domain unmanned cluster according to the given cross-domain unmanned cluster, longitudinally analyze the multi-layer coupling failure mechanism of the cross-domain unmanned cluster according to the connectivity between unmanned platform nodes in different spatial domains, and determine the resilience strategy for resilience process modeling. Transversely consider multiple time periods during the task execution process, divide the resilience process of the cross-domain unmanned cluster into a prevention stage, a degradation stage, a recovery stage, and a reconstruction stage, construct the resilience process model of the cross-domain unmanned cluster, and respectively construct the resilience indicators corresponding to each stage. Construct a multi-stage comprehensive resilience model of the cross-domain unmanned cluster according to the resilience indicators of the prevention stage, degradation stage, recovery stage, and reconstruction stage. Adopt the multi-stage comprehensive resilience model of the cross-domain unmanned cluster, and use the number of kill chains as a performance indicator to evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster to be evaluated under different scenarios.

2. The cross-domain unmanned cluster multi-stage comprehensive resilience evaluation method according to claim 1, wherein The resilience strategy includes: a replenishment and repair strategy, an intra-cluster reconstruction strategy, and an inter-cluster reconstruction strategy; the replenishment and repair strategy is used in the recovery stage of the multi-stage resilience process, and the intra-cluster reconstruction strategy and the inter-cluster reconstruction strategy are used in the reconstruction stage of the multi-stage resilience process. The replenishment and repair strategy is used to ensure that the system can use internal resources to replenish a certain unmanned platform when the supplies of the unmanned platform are insufficient, and if the unmanned platform suffers a minor failure, the system can repair the unmanned platform to ensure the stable and safe operation of the cross-domain intelligent unmanned cluster. The intra-cluster reconstruction strategy is used when a certain unmanned platform in the cross-domain unmanned cluster is deliberately damaged and fails or suffers a natural random failure, and the similar unmanned platforms in the same cluster as the failed unmanned platform can replace the failed unmanned platform to work; among them, the connection of the communication link between the new unmanned platform and the unmanned platform connected to the failed node is involved in the alternative work process. The inter-cluster reconstruction strategy is used when a certain unmanned platform in the cross-domain unmanned cluster is deliberately damaged and fails or suffers a natural random failure, and the similar unmanned platforms in other unmanned clusters can replace the failed unmanned platform to execute tasks; among them, the connection of the communication link between the new unmanned platform and the unmanned platform connected to the failed node is involved in the alternative work process.

3. The cross-domain unmanned cluster multi-stage comprehensive resilience evaluation method according to claim 1, characterized in that, The preventability indicator corresponding to the prevention stage is: Among them, is the preventable index of the cross-domain unmanned cluster, is the risk prediction ability, is the early warning ability, is the pre-event response ability.

4. The cross-domain unmanned cluster multi-stage comprehensive resilience assessment method according to claim 1, wherein The robustness indicator corresponding to the degradation stage is: Among them, is the robustness index of the cross-domain unmanned cluster, is the resistance coefficient, and the resistance coefficient is the ratio of the number of normal devices in the cross-domain unmanned cluster before and after resisting interference during the degradation stage, is the redundancy coefficient based on the task baseline, and the redundancy coefficient is the ratio of the part where the performance of the cross-domain unmanned cluster exceeds the task baseline to the task baseline, is the traditional resilience during the degradation stage.

5. The cross-domain unmanned cluster multi-stage comprehensive resilience evaluation method according to claim 1, characterized in that The recoverability indicator corresponding to the recovery stage is: Among them, is the coefficient of restitution, is the traditional toughness value in the recovery stage, is the value of the function of performance changing with time at moment, is the task baseline.

6. The cross-domain unmanned cluster multi-stage comprehensive resilience assessment method according to claim 1, wherein The reconfigurability indicator corresponding to the reconstruction stage is: Among them, is the reconfigurability index corresponding to the reconstruction stage, is the reconstruction coefficient, is the performance endpoint of cross-domain unmanned cluster reconstruction, is the initial performance of the cross-domain unmanned cluster, is the value of the function of performance varying with time at moment, is the task baseline.

7. The cross-domain unmanned cluster multi-stage comprehensive resilience evaluation method according to claim 3, characterized in that Construct a multi-stage comprehensive resilience model of the cross-domain unmanned cluster according to the resilience indicators of the prevention stage, degradation stage, recovery stage, and reconstruction stage as: Among them, is the multi-stage comprehensive resilience, and its value range is [0, 1]; , , , are the preventability index, robustness index, recoverability index, and reconfigurability index of the cross-domain unmanned cluster respectively, and their value ranges are all [0, 1].

8. The cross-domain unmanned cluster multi-stage comprehensive resilience assessment method according to claim 1, characterized in that The failure modes of the cross-domain unmanned cluster include: deliberate damage failure, natural random failure, and mixed failure. The deliberate damage failure includes maximum-degree damage and random damage. The natural random failure refers to equipment aging, component wear and looseness. The mixed failure is a failure mode under the combined action of natural random failure and deliberate damage failure.

9. A cross-domain unmanned cluster multi-stage comprehensive resilience assessment device, characterized in that, The device includes: The multi-layer coupling failure mechanism analysis module is used to construct a multi-layer heterogeneous network of the cross-domain unmanned cluster according to the given cross-domain unmanned cluster, longitudinally analyze the multi-layer coupling failure mechanism of the cross-domain unmanned cluster based on the connectivity between unmanned platform nodes in different spatial domains, and determine a resilience strategy for resilience process modeling; The resilience index determination module is used to horizontally consider multiple time periods during the task execution process, divide the resilience process of the cross-domain unmanned cluster into a prevention stage, a degradation stage, a recovery stage, and a reconstruction stage, construct a resilience process model of the cross-domain unmanned cluster, and respectively construct resilience indexes corresponding to each stage; The cross-domain unmanned cluster multi-stage comprehensive resilience model construction module is used to construct a multi-stage comprehensive resilience model of the cross-domain unmanned cluster according to the resilience indexes of the prevention stage, the degradation stage, the recovery stage, and the reconstruction stage; The cross-domain unmanned cluster multi-stage comprehensive resilience evaluation module is used to adopt the multi-stage comprehensive resilience model of the cross-domain unmanned cluster, take the number of kill chains as a performance index, and evaluate the multi-stage comprehensive resilience of the cross-domain unmanned cluster to be evaluated in different scenarios.

10. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the cross-domain unmanned cluster multi-stage comprehensive resilience evaluation method according to any one of claims 1 to 8.

Images