Intelligent unmanned cluster resilience measurement and evaluation method and device for autonomous recovery

Abstract

The application discloses a smart unmanned cluster resilience measurement and evaluation method and device for autonomous recovery, relates to the technical field of reliability and resilience evaluation, and is characterized in that the method is executed by a resilience evaluation device, and the resilience evaluation device can be deployed on a ground station, a cluster control node, an edge computing node or a simulation evaluation platform. A time-related fusion multi-damage variable resilience measurement model is constructed, and unified quantitative evaluation of the comprehensive performance of the smart unmanned cluster in maintaining task execution and autonomous recovery under external disturbance or internal failure is realized.

Description

Technical Field

[0001] This invention relates to the field of reliability and resilience evaluation technology, specifically to a method and apparatus for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery. Background Technology

[0002] During the long-term autonomous operation and task execution of intelligent unmanned swarms, the system often operates in an open, dynamic, and uncertain environment, continuously facing performance degradation caused by external disturbances and internal failures. Especially in applications focused on autonomous recovery, intelligent unmanned swarms must not only maintain basic task capabilities when disturbances occur, but also achieve continuous recovery and stable maintenance of task execution capabilities through mechanisms such as adaptive reconfiguration, resource reallocation, and capability repair. Therefore, how to calculably, comparablely, and generalize the resilience measurement of intelligent unmanned swarms throughout the entire "disturbance-degradation-recovery" process becomes a key foundation for supporting the resilience design and recovery decisions of intelligent unmanned swarms.

[0003] However, while some progress has been made in the research on resilience measurement of intelligent unmanned swarms, there are still significant shortcomings: First, existing resilience indicators have not yet formed a unified and universal measurement form, and are often constructed based on specific tasks or scenarios, making it difficult to make horizontal comparisons between different studies and to provide a consistent evaluation basis for general unmanned swarms; Second, existing methods mostly use static performance characteristics to characterize the absorption and recovery capabilities of the swarm, tending to use single-moment or a few discrete indicators to reflect the resilience level, making it difficult to depict the real-time performance changes of intelligent unmanned swarms over time during the task cycle, and thus failing to accurately reflect the dynamic recovery process after damage occurs; Third, existing measurements usually focus on a single dimension, lacking a unified integration of multiple factors such as task execution capability loss, resource stock consumption, and maintenance / recovery costs, making it difficult to comprehensively reveal the comprehensive trade-off between the ability of intelligent unmanned swarms to maintain task progress under dynamic disturbances and the resource costs incurred. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method and apparatus for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery.

[0005] This invention provides a method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery, including: In autonomous recovery scenarios, the task set, initial state variables, and non-destructive physical upper bound benchmark sequence obtained from benchmark simulation sequence under undamaged conditions are acquired for intelligent unmanned clusters. Obtain the moment when the intelligent unmanned swarm suffers damage during the execution of a set of tasks, and obtain the state variables and the physical upper bound benchmark of no damage corresponding to the current evaluation moment after the moment of damage that requires a resilience evaluation. Using the state variables and non-destructive physical upper bound benchmarks corresponding to the current evaluation time, the task progress damage variables, task execution resource coordination margin damage variables, individual state equilibrium damage variables, and shared maintenance reserve damage variables corresponding to the intelligent unmanned swarm at the current evaluation time are calculated respectively. Using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable corresponding to the intelligent unmanned cluster at the current evaluation time, the damage-induced state potential energy of the intelligent unmanned cluster at the current evaluation time is calculated. Based on the potential energy of the destructive state at the current evaluation moment, the dynamic resilience value of the intelligent unmanned swarm at the current evaluation moment is calculated, and the resilience of the intelligent unmanned swarm is measured and evaluated using the dynamic resilience value of the intelligent unmanned swarm at the current evaluation moment.

[0006] Preferably, the initial state variable includes each task in the task set. Task demand Task Importance Unmanned equipment initial energy resources and initial task capability resources Initial shared maintenance resources with intelligent unmanned swarms ; The non-destructive physical upper bound reference sequence includes: Included in every moment Upper bound of task completion rate under non-destructive conditions The upper bound of the task completion rate benchmark sequence; Included in every moment Upper bound of resource coordination margin for task execution at the same time under non-destructive conditions The upper bound benchmark sequence for task execution resource coordination margin; Included in every moment Upper bound of individual state equilibrium at the same time under non-destructive conditions The upper bound benchmark sequence of individual state equilibrium; Included in every moment Upper bound of shared maintenance reserve under non-destructive conditions The shared maintenance reserve upper limit benchmark sequence.

[0007] Preferably, using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each task in the initial state variables Task demand Task Importance and the current evaluation time The amount of completed tasks in the corresponding state variables Computational intelligent unmanned swarm at the current evaluation moment Weighted task completion rate ; Based on the intelligent unmanned swarm at the current evaluation time Weighted task completion rate and the current evaluation moment Upper bound of task completion rate under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding task progress damage variable .

[0008] Preferably, using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each unmanned equipment in the initial state variables initial energy resources and initial task capability resources and the current evaluation time Each unmanned equipment in the corresponding state variables Remaining energy resources and remaining task capabilities and resources Calculate the intelligent unmanned swarm at the current evaluation time. Energy resource retention rate and task capability resource retention rate ; Based on the intelligent unmanned swarm at the current evaluation time Energy resource retention rate and task capability resource retention rate Constructing intelligent unmanned swarms at the current evaluation moment Task execution resource coordination margin ; Based on the intelligent unmanned swarm at the current evaluation time Task execution resource coordination margin and the current evaluation time Upper bound of resource coordination margin for task execution at the same time under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding task execution resource coordination margin impairment variable .

[0009] Preferably, using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each unmanned equipment in the initial state variables initial energy resources and initial task capability resources and the current evaluation time Each unmanned equipment in the corresponding state variables Remaining energy resources and remaining task capabilities and resources Calculate the values ​​of each unmanned device in the intelligent unmanned swarm. At the current evaluation moment Normalized energy resource retention rate and normalized task capability resource retention rate ; Utilizing each unmanned device in the intelligent unmanned cluster At the current evaluation moment Normalized energy resource retention rate Calculate at the current evaluation time The energy resource balance factor, while utilizing each unmanned equipment in the intelligent unmanned cluster. At the current evaluation moment Normalized task capability resource retention rate Calculate at the current evaluation time Task capability resource balance factor; N represents normalization; Based on the current evaluation time The energy resource balance factor and the task capacity resource balance factor are constructed at the current evaluation time. The individual state balance, and based on the current evaluation time. The individual state balance and the current evaluation time Upper bound of individual state equilibrium at the same time under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding individual state equilibrium damage variable .

[0010] Preferably, using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on the initial shared maintenance resources of the intelligent unmanned cluster in the initial state variables. and the current evaluation moment The corresponding state variables include the shared maintenance surplus resources of the intelligent unmanned cluster. Computational intelligent unmanned swarm at the current evaluation moment Cluster remaining shared maintenance resource retention rate ; Based on the intelligent unmanned swarm at the current evaluation time Cluster remaining shared maintenance resource retention rate and the current evaluation moment Upper bound of shared maintenance reserves under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding shared maintenance reserve damage variable .

[0011] Preferably, the damage-induced state potential energy of the intelligent unmanned swarm at the current evaluation moment is calculated using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable corresponding to the intelligent unmanned swarm at the current evaluation moment, including: in, For intelligent unmanned swarms at the current evaluation moment The destructive induced state potential energy, The scaling factor for the potential energy is denoted as . The sensitivity coefficient for task progress impairment. The sensitivity coefficient for resource impairment during task execution. The individual state equilibrium damage sensitivity coefficient, To share the damage sensitivity coefficient for maintenance reserves, and .

[0012] This invention also provides a device for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery. The device is used to implement the method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery, and includes: The acquisition module is used to acquire, in the context of autonomous recovery scenarios, the task set, initial state variables, and non-destructive physical upper bound benchmark sequence obtained from the benchmark simulation sequence under the condition of no destruction; acquire the time of destruction when the intelligent unmanned cluster suffers destruction during the execution of the task set; and acquire the state variables and non-destructive physical upper bound benchmark corresponding to the current evaluation time after the time of destruction when resilience evaluation is required. The calculation module is used to calculate, using the state variables corresponding to the current evaluation time and the non-destructive physical upper bound benchmark, the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable of the intelligent unmanned swarm at the current evaluation time; and to calculate the destruction-induced state potential energy of the intelligent unmanned swarm at the current evaluation time using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable of the intelligent unmanned swarm at the current evaluation time. The resilience measurement and evaluation module is used to calculate the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time based on the potential energy of the destruction-induced state at the current evaluation time, and to perform resilience measurement and evaluation processing on the intelligent unmanned cluster using the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time.

[0013] This invention also provides a computing device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the intelligent unmanned swarm resilience measurement and evaluation method for autonomous recovery.

[0014] This invention also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery.

[0015] The beneficial effects of the invention are that by constructing time-related resilience indicators, it overcomes the limitation of traditional static assessment methods in being unable to characterize dynamic recovery processes; the method is not dependent on specific platforms and application scenarios and is applicable to multiple types of intelligent unmanned swarms; at the same time, it integrates task execution capabilities and various resource consumption into the resilience assessment framework, providing a quantifiable, comparable, and scalable theoretical basis for the resilience design and autonomous recovery decision-making of intelligent unmanned swarms. Attached Figure Description

[0016] Figure 1This is a flowchart of a method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery, provided by the present invention. Figure 2 This is a flowchart of the intelligent unmanned swarm resilience measurement and evaluation method for autonomous recovery provided by the present invention; Figure 3 This is a schematic diagram of an intelligent unmanned swarm resilience measurement and evaluation device for autonomous recovery provided by the present invention. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, suffixes such as "module," "part," or "unit" used to denote elements are used only for the purpose of illustrative purposes and have no inherent meaning. Therefore, "module," "part," or "unit" may be used interchangeably.

[0018] Figure 1 This is a flowchart of a method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery, provided by the present invention. Figure 1 As shown, it includes: Step S101: In the scenario of autonomous recovery, obtain the task set, initial state variables, and non-destructive physical upper bound benchmark sequence obtained from the benchmark simulation sequence under the condition of no destruction of the intelligent unmanned swarm; Furthermore, the initial state variable includes each task in the task set. Task demand Task Importance Unmanned equipment initial energy resources and initial task capability resources Initial shared maintenance resources with intelligent unmanned swarms ; Furthermore, the fundamental constraint is: for any , and , , , , .

[0019] The non-destructive physical upper bound reference sequence includes: Included in every moment Upper bound of task completion rate under non-destructive conditions The upper bound of the task completion rate benchmark sequence; Included in every moment Upper bound of resource coordination margin for task execution at the same time under non-destructive conditions The upper bound benchmark sequence for task execution resource coordination margin; Included in every moment Upper bound of individual state equilibrium at the same time under non-destructive conditions The upper bound benchmark sequence of individual state equilibrium; Included in every moment Upper bound of shared maintenance reserve under non-destructive conditions The shared maintenance reserve upper limit benchmark sequence.

[0020] Step S102: Obtain the time of damage that the intelligent unmanned cluster suffers during the execution of the task set, and obtain the state variables and the physical upper bound benchmark of no damage corresponding to the current evaluation time after the time of damage that requires resilience evaluation. It should be noted that the moment of damage and the current evaluation moment are not the same moment. There is only one moment when damage occurs, while the current evaluation moment is every moment after the damage occurs that requires evaluation.

[0021] Step S103: Using the state variables corresponding to the current evaluation time and the non-destructive physical upper bound benchmark, calculate the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable and shared maintenance reserve damage variable corresponding to the intelligent unmanned cluster at the current evaluation time. In one specific embodiment of this application, using the state variable corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each task in the initial state variables Task demand Task Importance and the current evaluation time The amount of completed tasks in the corresponding state variables The computational intelligent unmanned swarm at the current evaluation moment Weighted task completion rate ; Based on the intelligent unmanned swarm at the current evaluation time Weighted task completion rate and the current evaluation moment Upper bound of task completion rate under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding task progress damage variable .

[0022] In one specific embodiment of this application, using the state variable corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each unmanned equipment in the initial state variables initial energy resources and initial task capability resources and the current evaluation time Each unmanned equipment in the corresponding state variables Remaining energy resources and remaining task capabilities and resources Calculate the intelligent unmanned swarm at the current evaluation time. Energy resource retention rate and task capability resource retention rate ; Based on the intelligent unmanned swarm at the current evaluation time Energy resource retention rate and task capability resource retention rate Constructing intelligent unmanned swarms at the current evaluation moment Task execution resource coordination margin ; Based on the intelligent unmanned swarm at the current evaluation time Task execution resource coordination margin and the current evaluation time Upper bound of resource coordination margin for task execution at the same time under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding task execution resource coordination margin impairment variable .

[0023] In one specific embodiment of this application, using the state variable corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each unmanned equipment in the initial state variables initial energy resources and initial task capability resources and the current evaluation time Each unmanned equipment in the corresponding state variables Remaining energy resources and remaining task capabilities and resources Calculate the values ​​of each unmanned device in the intelligent unmanned swarm. At the current evaluation moment Normalized energy resource retention rate and normalized task capability resource retention rate Where N represents normalization; Utilizing each unmanned device in the intelligent unmanned cluster At the current evaluation moment Normalized energy resource retention rate Calculate at the current evaluation time The energy resource balance factor, while utilizing each unmanned equipment in the intelligent unmanned cluster. At the current evaluation moment Normalized task capability resource retention rate Calculate at the current evaluation time Task capability resource balance factor; Based on the current evaluation time The energy resource balance factor and the task capacity resource balance factor are constructed at the current evaluation time. The individual state balance, and based on the current evaluation time. The individual state balance and the current evaluation time Upper bound of individual state equilibrium at the same time under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding individual state equilibrium damage variable .

[0024] In one specific embodiment of this application, using the state variable corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on the initial shared maintenance resources of the intelligent unmanned cluster in the initial state variables. and the current evaluation moment The corresponding state variables include the shared maintenance surplus resources of the intelligent unmanned cluster. The computational intelligent unmanned swarm at the current evaluation moment Cluster remaining shared maintenance resource retention rate ; Based on the intelligent unmanned swarm at the current evaluation time Cluster remaining shared maintenance resource retention rate and the current evaluation moment Upper bound of shared maintenance reserves under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding shared maintenance reserve damage variable .

[0025] Step S104: Calculate the damage-induced state potential energy of the intelligent unmanned cluster at the current evaluation time using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable corresponding to the intelligent unmanned cluster at the current evaluation time. In one specific embodiment of this application, the damage-induced state potential energy of the intelligent unmanned swarm at the current evaluation moment is calculated using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable corresponding to the current evaluation moment. in, For intelligent unmanned swarms at the current evaluation moment The destructive induced state potential energy, The scaling factor for the potential energy is denoted as . The sensitivity coefficient for task progress impairment. The sensitivity coefficient for resource impairment during task execution. The individual state equilibrium damage sensitivity coefficient, To share the damage sensitivity coefficient for maintenance reserves, and .

[0026] Furthermore, , , All are less than or equal to 1; when any variable reaches 1, it is treated as U(t) tends to infinity and R(t) tends to 0.

[0027] Step S105: Based on the potential energy of the destructive state of the intelligent unmanned cluster at the current evaluation time, calculate the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time, and use the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time to perform resilience measurement and evaluation processing on the intelligent unmanned cluster.

[0028] This invention provides a method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery. This method is executed by a resilience evaluation device, which can be deployed at a ground station, swarm control node, edge computing node, or simulation evaluation platform. It constructs a time-dependent resilience measurement model that integrates multiple damage variables, enabling a unified quantitative assessment of the comprehensive performance of intelligent unmanned swarms in maintaining mission execution and autonomously recovering under external disturbances or internal failures. The method's inputs include the evaluation object, state variables, and a non-destructive physical upper bound benchmark; the outputs include state potential energy at each time step, dynamic resilience values, and resilience statistics under multiple simulation conditions, such as... Figure 2 As shown, the specific steps include: Step S100: Input the evaluation object, state variables, and non-destructive physical upper bound benchmark; The evaluation object is the intelligent unmanned swarm system, which includes the intelligent unmanned swarm and the task set.

[0029] Suppose that the intelligent unmanned swarm includes One unmanned device, denoted as: Suppose the task set contains One task, denoted as: State variables are represented as follows: Task The task requirement is The task importance is ,time Completed task quantity is Unmanned equipment At any moment The remaining energy resources are The initial energy resources are The remaining task capacity resources are The initial task capability resources are Cluster-shared maintenance resources are available at all times. The remaining amount is Initial shared maintenance resources are .

[0030] Furthermore, the fundamental constraint is: for any , and , , , , .

[0031] Under the same task, the same initial resources, and no-destructive conditions, a physical upper bound benchmark is given. This benchmark is an attainable upper bound of the state given by the task execution model or benchmark simulation sequence under no-destructive conditions.

[0032] There are four types of non-destructive physical upper bounds, which are represented as follows. Indicates time under non-destructive conditions The upper bound of the task completion rate; This represents the upper bound of resource coordination margin for task execution at the same time under non-destructive conditions. It represents the upper bound of the equilibrium of individual states at the same moment under non-destructive conditions; This represents the upper bound of the shared maintenance reserve. Since maintenance resources are shared reserves and are not accessed unless damaged, this value is taken as... .

[0033] Step S200: Calculate the task progress impairment variables; The resilience evaluation device calculates the time based on the current completion amount of each task. Cluster weighted task completion rate : in, The range of values ​​is , indicating time The percentage of weighted tasks that have been completed in the cluster.

[0034] To reflect the damage to mission progress caused by destruction, time is based on the non-destructive condition. Upper bound of task completion rate Define the task progress impairment variable: because This is the physical upper bound of the task completion rate at any given moment under non-destructive conditions. Characterizes the relative damage to the actual mission propulsion relative to the non-destructive upper bound, with a value range of [value range missing]. .

[0035] Step S300: Calculate the task execution resource coordination margin impairment variable and the individual state equilibrium impairment variable; Energy resources and mission capability resources are both individual mission execution resources distributed across various unmanned equipment. The resilience assessment device calculates the energy resource retention rate. and task capability resource retention rate : To reflect the bottleneck synergy relationship between energy resources and mission capability resources, a weighted harmonic average is used to construct the mission execution resource synergy margin. : in, and The weights for energy resources and mission capability resources are respectively, satisfying: set up To define the upper bound of task execution resource coordination margin at the same time in the non-destructive physical upper bound benchmark, define the task execution resource coordination margin impairment variable. : in, Characterizes the degree of damage to the current task execution resource coordination margin relative to the undamaged physical upper bound after the destruction and recovery process, with a value range of [value missing]. .

[0036] Given the same total amount of resources, the more evenly resources are distributed among the unmanned equipment, the stronger the cluster's resilience against subsequent damage and mission degradation. For any resource type... Define unmanned equipment Normalized resource retention rate : in, For unmanned equipment initial energy resources Or initial task capability resources ; For unmanned equipment At any moment energy resources or task ability resources .

[0037] Define the mean normalized retention rate of this type of resource. : Define the dispersion of this type of resource distribution : Define this type of resource balance factor : The resource balance factor includes two types: energy resources and mission capability resources.

[0038] Energy resource balance factor and task capability resource balance factor Constructing individual state equilibrium : in, and The weights for energy resource balance and task capacity resource balance are respectively, satisfying: set up Define the individual state equilibrium damage variable as the upper bound of the individual state equilibrium at the same moment in the non-destructive physical upper bound benchmark. : Step S400: Calculate the shared maintenance reserve damage variable; Assume the remaining shared maintenance resource retention rate of the cluster. for: Let the upper bound of shared maintenance reserves be Take without damage Define shared maintenance reserve damage variables : in, Characterizes the degree of damage to the shared maintenance reserve relative to the upper bound of the non-destructive reserve, with a value range of [value range missing]. .

[0039] Step S500: Calculate the state potential energy and dynamic toughness.

[0040] The resilience evaluation device constructs a time-based model for intelligent unmanned swarms based on task progress damage variables, task execution resource damage variables, individual state equilibrium damage variables, and shared maintenance reserve damage variables. Destructive induced state potential energy : in, The scaling factor for the potential energy is denoted as . The sensitivity coefficient for task progress impairment. The sensitivity coefficient for resource impairment during task execution. The individual state equilibrium damage sensitivity coefficient, To share the damage sensitivity coefficient for maintenance reserves, and . , , All are less than or equal to 1; when any one of the variables reaches 1, it is calculated as follows: Approaching infinity Approaching 0 processing.

[0041] When no damage occurs or the system state recovers to the physical upper bound of no damage, there is ,but When damage leads to mission delays, reduced resource coordination margins, uneven distribution of individual statuses, or depletion of shared maintenance reserves, at least one damage variable increases. It increases accordingly.

[0042] Ultimately, the intelligent unmanned swarm is computed at time... Dynamic resilience Used to measure the resilience of intelligent unmanned swarms: because ,therefore The range of values ​​is This resilience index is determined solely by time intervals. The recovery strategy is calculated based on the task completion status, energy status, task capability status, shared maintenance resource status, and their non-destructive physical upper bounds, and does not directly depend on the specific maintenance strategy, task reallocation strategy, or task sequence adjustment strategy adopted. The recovery strategy only indirectly affects the state variables at subsequent time steps. .

[0043] When considering statistical resilience under random failure conditions, the above steps are repeated using a Monte Carlo experiment. Next, get the first The simulation is at time [time]. toughness value And calculate the average toughness This is used to evaluate the resilience of intelligent unmanned swarms: (twenty one) Compared with existing technologies, this invention has at least the following advantages: by constructing time-related resilience indicators, it overcomes the limitation of traditional static evaluation methods in being unable to characterize dynamic recovery processes; the method is not dependent on specific platforms and application scenarios and is applicable to multiple types of intelligent unmanned swarms; at the same time, it integrates task execution capabilities and multiple types of resource consumption into the resilience evaluation framework, providing a quantifiable, comparable, and scalable theoretical basis for the resilience design and autonomous recovery decision-making of intelligent unmanned swarms.

[0044] Figure 3 This is a schematic diagram of an intelligent unmanned swarm resilience measurement and evaluation device for autonomous recovery provided by the present invention, as shown below. Figure 3 As shown, the device is used to implement the intelligent unmanned swarm resilience measurement and evaluation method for autonomous recovery, including: The acquisition module is used to acquire, in the context of autonomous recovery scenarios, the task set, initial state variables, and non-destructive physical upper bound benchmark sequence obtained from the benchmark simulation sequence under the condition of no destruction; acquire the time of destruction when the intelligent unmanned cluster suffers destruction during the execution of the task set; and acquire the state variables and non-destructive physical upper bound benchmark corresponding to the current evaluation time after the time of destruction when resilience evaluation is required. The calculation module is used to calculate, using the state variables corresponding to the current evaluation time and the non-destructive physical upper bound benchmark, the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable of the intelligent unmanned swarm at the current evaluation time; and to calculate the destruction-induced state potential energy of the intelligent unmanned swarm at the current evaluation time using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable of the intelligent unmanned swarm at the current evaluation time. The resilience measurement and evaluation module is used to calculate the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time based on the potential energy of the destruction-induced state at the current evaluation time, and to perform resilience measurement and evaluation processing on the intelligent unmanned cluster using the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time.

[0045] The present invention also provides a computing device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the intelligent unmanned swarm resilience measurement and evaluation method for autonomous recovery.

[0046] The present invention also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery.

[0047] The invention will now be illustrated using a case study of a drone swarm urban ground information station data collection task. This embodiment is used to illustrate the variable calculation relationships of the invention and does not limit the scope of protection of the invention.

[0048] Step S100: Establish scene parameters and non-destructive physical upper bound benchmarks; Assume an intelligent unmanned swarm consisting of 3 UAVs, performing information collection tasks for 8 ground information stations. The initial parameters of the UAVs are shown in Table 1, where shared maintenance resources are considered. For unified storage across the cluster, tasks are not assigned to individual drones. Task parameters are shown in Table 2.

[0049] Table 1 Initial Parameter Settings for Unmanned Aerial Vehicle Cluster In this embodiment, the initial amount of shared maintenance resources is set as follows: .

[0050] Table 2 Initial Parameter Settings for the Task The evaluation parameters are shown in Table 3. The non-destructive physical upper bound is obtained from the simulation of the task execution without destruction; the destruction state is obtained from the instantaneous destruction and its subsequent state evolution under the same task scenario.

[0051] Table 3 Toughness Calculation Parameter Settings Steps S200 to S400: Calculate each damage variable; Select =5 min、 =13 min and =18 min. The calculation process is explained at three time points. The resilience evaluation device obtains the state variables and damage variables at each level based on the task completion amount, energy resources, task capability resources, shared maintenance resources, and the non-destructive physical upper limit benchmark, as shown in Table 4.

[0052] Table 4. Calculation results of state variables and damage variables at each time point. by Taking 5 min as an example, the task progress loss is calculated by equation (4): Task execution resource damage is calculated by equation (9): Individual state equilibrium damage is calculated by equation (16): Shared maintenance reserve damage is calculated using equation (18): Step S500: Calculate the state potential energy and dynamic toughness.

[0053] Substituting the damage variables in Table 4 into Equations (19) and (20), we obtain the state potential energy and dynamic toughness values ​​at each time point, as shown in Table 5.

[0054] Table 5. Calculation results of potential energy and dynamic toughness. by Taking 5 min as an example, based on the parameter values ​​in Table 3 and the damage variables in Table 4, the state potential energy is: The dynamic toughness is obtained from equation (20): As shown in Table 5, dynamic resilience decreases after the damage occurs; as the state recovers, the damage to task progress and the damage to individual state equilibrium decrease, and dynamic resilience recovers. Since shared maintenance resources have been consumed, even if the task progress is gradually restored, the resilience value will not directly equal 1 because the task can eventually be completed, thus reflecting the true state of the intelligent unmanned swarm in resisting further task degradation at the current moment.

[0055] In the Monte Carlo simulation evaluation, the toughness evaluation device repeatedly performed the above steps for different failure times, different damaged unmanned equipment, and different damage degrees, obtaining multiple sets of results. And calculate the average resilience sequence according to Equation (21) to compare the state resilience performance under different cluster configurations or different recovery schemes.

[0056] The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, but this does not limit the scope of the invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and spirit of the present invention should be within the scope of the present invention.

Claims

1. A method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery, characterized in that, include: In autonomous recovery scenarios, the task set, initial state variables, and non-destructive physical upper bound benchmark sequence obtained from benchmark simulation sequence under undamaged conditions are acquired for intelligent unmanned clusters. Obtain the moment when the intelligent unmanned swarm suffers damage during the execution of a set of tasks, and obtain the state variables and the physical upper bound benchmark of no damage corresponding to the current evaluation moment after the moment of damage that requires a resilience evaluation. Using the state variables and non-destructive physical upper bound benchmarks corresponding to the current evaluation time, the task progress damage variables, task execution resource coordination margin damage variables, individual state equilibrium damage variables, and shared maintenance reserve damage variables corresponding to the intelligent unmanned swarm at the current evaluation time are calculated respectively. Using the task progress impairment variables, task execution resource coordination margin impairment variables, individual state equilibrium impairment variables, and shared maintenance reserve impairment variables corresponding to the intelligent unmanned swarm at the current evaluation moment, the damage-induced state potential energy of the intelligent unmanned swarm at the current evaluation moment is calculated, which includes: in, For intelligent unmanned swarms at the current evaluation moment The destructive induced state potential energy, The scaling factor for the potential energy is denoted as . The sensitivity coefficient for task progress impairment. The sensitivity coefficient for resource impairment during task execution. The individual state equilibrium damage sensitivity coefficient, To share the damage sensitivity coefficient for maintenance reserves, and ; Based on the potential energy of the destructive state at the current evaluation moment, the dynamic resilience value of the intelligent unmanned swarm at the current evaluation moment is calculated, and the resilience of the intelligent unmanned swarm is measured and evaluated using the dynamic resilience value of the intelligent unmanned swarm at the current evaluation moment. The initial state variable includes each task in the task set. Task demand Task Importance Unmanned equipment initial energy resources and initial task capability resources Initial shared maintenance resources with intelligent unmanned swarms ; The non-destructive physical upper bound reference sequence includes: Included in every moment Upper bound of task completion rate under non-destructive conditions The upper bound of the task completion rate benchmark sequence; Included in every moment Upper bound of resource coordination margin for task execution at the same time under non-destructive conditions The upper bound benchmark sequence for task execution resource coordination margin; Included in every moment Upper bound of individual state equilibrium at the same time under non-destructive conditions The upper bound benchmark sequence of individual state equilibrium; Included in every moment Upper bound of shared maintenance reserve under non-destructive conditions The shared maintenance reserve upper limit benchmark sequence.

2. The method according to claim 1, characterized in that, Using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each task in the initial state variables Task demand Task Importance and the current evaluation time The amount of completed tasks in the corresponding state variables Computational intelligent unmanned swarm at the current evaluation moment Weighted task completion rate ; Based on the intelligent unmanned swarm at the current evaluation time Weighted task completion rate and the current evaluation moment Upper bound of task completion rate under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding task progress damage variable .

3. The method according to claim 2, characterized in that, Using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each unmanned equipment in the initial state variables initial energy resources and initial task capability resources and the current evaluation time Each unmanned equipment in the corresponding state variables Remaining energy resources and remaining task capabilities and resources Calculate the intelligent unmanned swarm at the current evaluation time. Energy resource retention rate and task capability resource retention rate ; Based on the intelligent unmanned swarm at the current evaluation time Energy resource retention rate and task capability resource retention rate Constructing intelligent unmanned swarms at the current evaluation moment Task execution resource coordination margin ; Based on the intelligent unmanned swarm at the current evaluation time Task execution resource coordination margin and the current evaluation time Upper bound of resource coordination margin for task execution at the same time under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding task execution resource coordination margin impairment variable .

4. The method according to claim 3, characterized in that, Using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on each unmanned equipment in the initial state variables initial energy resources and initial task capability resources and the current evaluation time Each unmanned equipment in the corresponding state variables Remaining energy resources and remaining task capabilities and resources Calculate the values ​​of each unmanned device in the intelligent unmanned swarm. At the current evaluation moment Normalized energy resource retention rate and normalized task capability resource retention rate N represents normalization; Utilizing each unmanned device in the intelligent unmanned cluster At the current evaluation moment Normalized energy resource retention rate Calculate at the current evaluation time The energy resource balance factor, while utilizing each unmanned equipment in the intelligent unmanned cluster. At the current evaluation moment Normalized task capability resource retention rate Calculate at the current evaluation time Task capability resource balance factor; Based on the current evaluation time The energy resource balance factor and the task capacity resource balance factor are constructed at the current evaluation time. The individual state balance, and based on the current evaluation time. The individual state balance and the current evaluation time Upper bound of individual state equilibrium at the same time under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding individual state equilibrium damage variable .

5. The method according to claim 4, characterized in that, Using the state variables corresponding to the current evaluation moment and the non-destructive physical upper bound benchmark, the following variables are calculated for the intelligent unmanned swarm at the current evaluation moment: task progress impairment variable, task execution resource coordination margin impairment variable, individual state equilibrium impairment variable, and shared maintenance reserve impairment variable: Based on the initial shared maintenance resources of the intelligent unmanned cluster in the initial state variables. and the current evaluation moment The corresponding state variables include the shared maintenance surplus resources of the intelligent unmanned cluster. Computational intelligent unmanned swarm at the current evaluation moment Cluster remaining shared maintenance resource retention rate ; Based on the intelligent unmanned swarm at the current evaluation time Cluster remaining shared maintenance resource retention rate and the current evaluation moment Upper bound of shared maintenance reserves under non-destructive conditions Calculate the intelligent unmanned swarm at the current evaluation time. The corresponding shared maintenance reserve damage variable .

6. A device for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery, the device being used to implement the method as described in any one of claims 1-5, characterized in that, include: The acquisition module is used to acquire, in the context of autonomous recovery scenarios, the task set, initial state variables, and non-destructive physical upper bound benchmark sequence obtained from the benchmark simulation sequence under the condition of no destruction; acquire the time of destruction when the intelligent unmanned cluster suffers destruction during the execution of the task set; and acquire the state variables and non-destructive physical upper bound benchmark corresponding to the current evaluation time after the time of destruction when resilience evaluation is required. The calculation module is used to calculate, using the state variables corresponding to the current evaluation time and the non-destructive physical upper bound benchmark, the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable of the intelligent unmanned swarm at the current evaluation time; and to calculate the destruction-induced state potential energy of the intelligent unmanned swarm at the current evaluation time using the task progress damage variable, task execution resource coordination margin damage variable, individual state equilibrium damage variable, and shared maintenance reserve damage variable of the intelligent unmanned swarm at the current evaluation time. The resilience measurement and evaluation module is used to calculate the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time based on the potential energy of the destruction-induced state at the current evaluation time, and to perform resilience measurement and evaluation processing on the intelligent unmanned cluster using the dynamic resilience value of the intelligent unmanned cluster at the current evaluation time.

7. A computing device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the intelligent unmanned swarm resilience measurement and evaluation method for autonomous recovery as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method for measuring and evaluating the resilience of intelligent unmanned swarms for autonomous recovery as described in any one of claims 1 to 5.

Images