An importance-based equipment system task reliability optimization method

Abstract

The application discloses an equipment system task reliability optimization method based on importance, belongs to the technical field of equipment system and reliability optimization, and mainly solves the problems that static importance measurement is disengaged from a time-space dynamic environment in a traditional equipment system maintenance optimization strategy and that there is a lack of adaptive scheduling algorithm under resource restriction. The application firstly constructs a dynamic node importance index based on a task chain, quantifies the key degree of a node to a system task completion capability, secondly establishes a maintenance scheduling optimization model under the restriction of limited support resources, takes the maximization of task period reliability recovery as an objective, and proposes a dynamic repair scheduling algorithm considering time-space constraints, so that limited maintenance resources are always preferentially allocated to a key node with the largest current contribution; and simulation results show that the method can effectively improve the task reliability of the system under random failure, deliberate attack and regional attack scenes, and provides theoretical and technical support for guiding the maintenance scheduling and auxiliary decision of the equipment system in a high confrontation environment.

Description

Technical Field

[0001] This invention belongs to the field of equipment system and reliability optimization technology, and mainly relates to a method for optimizing the reliability of equipment system tasks based on importance. Background Technology

[0002] Unmanned Systems of Systems (USoS), as the core support for modern complex mission execution, are complex mega-systems formed by integrating multiple functionally heterogeneous, autonomously operating, and deeply collaborative subsystems through networked interfaces. Their core advantage lies in achieving overall efficiency far exceeding the sum of individual subsystem capabilities through emergent effects and synergistic linkages between subsystems, playing an irreplaceable role in critical areas such as emergency support. However, the highly dynamic and adversarial nature of modern mission environments means that USoS constantly face multi-source uncertain shocks such as fire strikes, electromagnetic interference, and Trojan virus intrusions. These shocks can directly cause dynamic failures of nodes within the system, leading to link interruptions, functional path breaks, and ultimately a rapid decline in system mission reliability. More importantly, multi-source shocks are coupled with spatiotemporal constraints. For example, a regional attack can instantly paralyze all nodes and links within a specific spatial range; dynamic electromagnetic interference can affect link connectivity in different regions over time; and node maneuvering can alter its interaction range with other nodes. These coupling effects further exacerbate the fluctuation and degradation risks of system mission capabilities.

[0003] Maintenance and optimization, as a key means to improve the mission reliability of equipment systems, aims to quickly restore the functionality of failed nodes and maintain the supply capacity of the system's effective mission chain under the premise of limited resources, through scientific repair and scheduling strategies. However, existing equipment system maintenance and optimization strategies still have many limitations that are difficult to overcome, and cannot adapt to the needs of dynamic spatiotemporal environments and high-confrontation mission scenarios: First, static importance metrics are disconnected from dynamic task requirements. Most existing optimization scheduling strategies rely on node importance indicators based on static topology structures, such as node degree, betweenness, and topology contribution rate. These indicators only reflect the inherent importance of nodes in a fixed network structure, but fail to fully capture the real time-varying contribution of nodes to task execution in a dynamic spatiotemporal environment. In actual task scenarios, the importance of nodes changes dynamically with their location, running state, task stage, and the failure status of other nodes. Static indicators cannot update these dynamic changes in real time, leading to biased repair priority judgments. This often results in resource waste due to a mismatch between repaired nodes and current task requirements, making it difficult to achieve efficient recovery of the equipment system's mission capabilities.

[0004] Secondly, there is a lack of adaptive scheduling algorithms under resource constraints. Support resources for equipment systems, such as the number of maintenance units and repair time, are typically subject to strict constraints. Most existing maintenance strategies employ predefined fixed scheduling rules or optimization schemes based on offline computation, lacking an adaptive scheduling mechanism capable of responding in real-time to battle damage status and dynamically adjusting repair priorities. When faced with sudden large-scale shocks, these strategies cannot quickly reconstruct the repair sequence, leading to long-term disruptions in the core mission chain, making it difficult for the reliability of the equipment system to recover quickly, and even resulting in permanent paralysis.

[0005] Third, there is a lack of deep coupling between maintenance decisions and mission chain effectiveness. The mission reliability of an equipment system is essentially determined by the quantity and quality of effective mission chains, and the core of maintenance optimization should be ensuring the continuous availability of critical mission chains. However, existing strategies often assess the importance of individual nodes in isolation, failing to directly link node repair with mission chain availability. For example, some strategies only focus on restoring the functionality of the node itself, neglecting whether the repaired node can quickly integrate into the current effective mission chain and fill critical functional gaps. This decision-making logic of "emphasizing node repair and neglecting mission chain adaptation" means that even if a large number of nodes are repaired, the equipment system may still fail to form an effective mission chain due to incomplete functional paths, resulting in poor mission capability recovery.

[0006] Fourth, there is insufficient adaptability to multi-source impacts. Most existing maintenance strategies employ generalized scheduling logic, failing to adapt to the different damage patterns of various impact types. For example, random attacks causing scattered node failures necessitate prioritizing the repair of nodes most critical to the task chain coverage; deliberate attacks targeting core nodes require rapid repair to reconstruct the system's core framework; regional attacks causing concentrated failures of local nodes require balancing regional functional complementarity with efficient allocation of maintenance resources. Existing strategies lack the ability to respond to these differentiated needs, resulting in difficulties achieving optimal repair results under various impact scenarios.

[0007] In summary, existing equipment system maintenance optimization strategies have significant shortcomings in dynamic importance assessment, adaptive scheduling under resource constraints, and the coupling of task chains and maintenance decisions, making it difficult to meet the demands for rapid recovery of system mission reliability in modern high-confrontation, spatiotemporally dynamic environments. Therefore, there is an urgent need for a method that can quantify dynamic importance based on the real-time contribution of nodes to task chain availability and achieve adaptive repair scheduling under resource-constrained conditions, providing support for the continuous mission completion capability of equipment systems in complex mission environments. Summary of the Invention

[0008] To overcome the shortcomings of existing technologies and address the problems of static importance measurement being disconnected from the spatiotemporal dynamic environment and the lack of adaptive scheduling algorithms under resource constraints in traditional equipment system maintenance optimization strategies, this paper proposes an importance-based equipment system task reliability optimization method. This method dynamically updates dynamic importance by quantifying the real-time contribution of nodes to the availability of the task chain, designs an adaptive repair scheduling mechanism in conjunction with resource constraints, and prioritizes the allocation of maintenance resources to critical failure nodes. This achieves the maximum recovery of equipment system task reliability under multi-source uncertain shocks, providing technical support for equipment system maintenance decisions and resource scheduling.

[0009] The technical solution adopted by this invention to solve its technical problem includes the following steps: Step 1: Construct a directed graph based on the basic functional units of the equipment system. With directed graph Based on this, a task chain is constructed, breaking down the target task into an ordered logical flow sequence; the ordered logical flow sequence includes... The logical flow of each stage; simulating and calculating the directed graph based on the ordered logical flow sequence and the node set in the directed graph. Middle node At any moment Dynamic importance ; Step 2: Construct a reliability recovery model; The reliability recovery model includes task reliability. and the reliability recovery objective function; The reliability of the construction task for: ; In the formula, Total number of Monte Carlo simulations; For the first Next calculation time The result of the total number of task chains in the directed graph; if the total number of task chains is greater than zero, then... The value is 1; if the total number of task chains is less than or equal to zero, then The value of is 1; The reliability recovery objective function is: ; in, The total known task duration; It is the set of instantaneous states of all nodes at all times; To maintain scheduling decisions; the maintenance of scheduling decisions To ensure that each available guarantee node C in the guarantee resource pool is at time [time missing] The process involves determining which failed node to allocate resources for repair, where the resource pool comprises all nodes of type "protected node"; the maintenance scheduling decision... The core maintenance constraint set must be satisfied; the core maintenance constraint set includes maintenance resource constraints and maintenance time constraints; the reliability recovery objective function is a single-objective optimization function. To represent the task reliability function, The value is determined by time. , and maintenance scheduling decisions To be determined jointly; Step 3: Real-time traversal of the nodes of the directed graph at time... The instantaneous state is used to identify the current fault state. And it is not in maintenance condition. The identified nodes are marked as candidate states to be repaired; the set of all nodes marked as candidate states to be repaired is the candidate node set. Real-time calculation of candidate node set The current dynamic importance of each node. In the candidate node set In the process, nodes are sorted from highest to lowest dynamic importance to obtain a dynamic priority list. ; Search the resource pool to see if there are currently any available maintenance units. And the set of candidate nodes If not empty, then follow the dynamic priority list. The repair units are assigned to the candidate node set in the following order. Middle node, number of currently available maintenance units or candidate node set If empty, no maintenance unit will be allocated; the resource pool consists of all nodes of the guaranteed node type; if the candidate node set... The middle node is being repaired by the guaranteed nodes in the resource pool, and a candidate node set is being developed. The state of the intermediate node is changed from the fault state. Immediately switch to maintenance mode Record the set of candidate nodes for node repair. After the repair time of the intermediate nodes meets the maintenance time constraint, the candidate node set is determined. The middle node status is updated to running state. The repaired set of candidate nodes The middle node is reconnected to the directed graph, and the directed graph is updated in real time. Use the reliability recovery objective function to assess task reliability. Optimize to improve task reliability To maximize the mission completion capability of the equipment system.

[0010] Furthermore, the construction of the directed graph The steps of task chain M are as follows: Step S1-1-1: Construct a directed graph based on the basic functional units of the equipment system. ; Based on mission requirements, deploy the basic functional units of a heterogeneous equipment system in space; using all basic functional units as nodes; traverse all nodes. compute nodes Position relative to all other nodes And Euclidean distance; if node If the Euclidean distances between the node and all other nodes simultaneously satisfy both the distance communication condition and the delay condition, then the node is activated. The edge between the node and the corresponding node; if the node If the Euclidean distance between a node and all other nodes does not satisfy the distance and delay conditions, then the node... There is no edge between the corresponding node and the edge; Construct a directed graph based on all nodes and active edges. for: ; in, A set of nodes; The set of all active edges; The set of attributes; the set of nodes includes all nodes; the set of edges For all edges The set; the edges The types include reconnaissance edges, decision-making edges, execution edges, and support edges; after construction, the directed graph is updated in real time. ; The attribute set A collection of all node attributes; node attributes for: ; in, For node identifiers; The node function types include reconnaissance node S, decision-making node D, execution node W, support node C, and target node T. For the node at time The instantaneous state of the node at time [time]. The instantaneous state includes the running state. Processing state Non-working status Fault state and maintenance status ; For the node at time The position vector; For the node at time The velocity vector; These are node capability parameters; node capability parameters include communication distance. and processing capacity ; Time window for nodes ; This refers to the node startup time; This is the end time of the node; The edge Represents a directed graph In the middle, when the node To the node Directed functional interaction paths between entities that satisfy spatiotemporal constraints; ; Step S1-1-2: Using a directed graph Build a task chain based on this; The task chain for: ; in For task chains, For task chain The 1 node In the task chain Middle node The next node, For task chain Middle node and nodes The directed edges between them, Indicates from arrive Functional dependencies or execution connections, The number of nodes in the task chain; the task chain It must also satisfy topological constraints, functional coverage constraints, task execution constraints, and availability constraints.

[0011] The target task is decomposed into an ordered logical flow sequence; the ordered logical flow sequence includes... The logical flow of each stage; simulating and calculating the directed graph based on the ordered logical flow sequence and the node set in the directed graph. Middle node At any moment Dynamic importance The steps are as follows: Step S1-2-1: Based on the ordered logical flow sequence The Middle Each stage of logical flow, from the node set Extraction stage node subset stage node subset For a set of nodes The functional types of the middle node and the first All nodes with the same stage logic flow; Step S1-2-2: Based on the subset of stage nodes Constructing an adjacency matrix ; Adjacency Matrix Obtained through the Hadamard product, i.e.: ; in, For link matrix; For filtering matrix; For Hadamard product; Link Matrix The elements in are ,if Represents a node and There are logical or physical links between them; Filtering matrix elements in for: ; in For filtering matrix The Line 1 Column elements, For indicator functions, if ,but ,like ,but , For Euclidean distance, This represents the maximum communication distance between nodes. For node processing latency, The time required for the task chain to complete its function. For the first The maximum allowable delay for each stage; Step S1-2-3: Calculate the adjacency matrix for all stages. Based on the adjacency matrix of all stages Calculate the complete path count matrix for: ; The complete path count matrix elements in Represents starting from the node To the destination node The number of all feasible paths that satisfy the task chain definition; Step S1-2-4: Based on the complete path counting matrix Calculate the total number of task chains ; Total number of task chains The sum of all elements in the path counting matrix:

[0012] in It is a column vector of all 1s. It is a row vector consisting entirely of 1s; Step S1-2-5: Simulate the operation of the computing nodes At any moment Dynamic importance Dynamic importance For nodes In a directed graph, the ratio of the number of task chains in the running state to the number of task chains in the fault state is calculated using the following formula: ; in, For nodes At any moment The dynamic importance, For nodes The reliability of the target task in a directed graph when it is in the running state. For nodes When in a fault state, the reliability of the target task in the directed graph. The total number of task chains when the directed graph is created. The number of failed task chains. Represents a node This represents the number of task chains during runtime. Represents a node This represents the number of task chains in a faulty state.

[0013] Furthermore, the maintenance resource constraint is: at any given time The number of nodes in maintenance mode at the same time cannot exceed the maximum number of maintenance tasks that can be performed. .

[0014] .

[0015] Furthermore, the maintenance time constraint is a node Each repair requires a non-zero repair time. If the node At the point of time The repair process begins, and the time it takes for it to return to running state is... satisfy: .

[0016] Furthermore, the real-time updated directed graph The steps are as follows: If a directed graph If, after any node is moved, the edges between the moved nodes satisfy the spatiotemporal constraints, then the corresponding edges between the nodes are activated, and the activated edges are added to the edge set. If the edges between nodes after the move do not satisfy the spatiotemporal constraints, then the edges will be removed from the edge set. Remove edges that do not satisfy the spatiotemporal constraints, thereby continuously updating the directed graph; Furthermore, the node and The Euclidean distance between them satisfies both the distance communication determination condition and the delay determination condition as follows: ; like ,but ,like ,but ; The determination result is to simultaneously satisfy both the distance communication determination condition and the delay determination condition; The distance communication determination condition is the node. and The communication distance between them is less than the maximum communication distance. ; The delay determination condition is that within the maximum communication distance, the node... and Transmission delay between them It must be below the maximum threshold allowed by the task. And the interaction must occur on the node. and Common runtime window between Inside.

[0017] The topological constraints, functional coverage constraints, task execution constraints, and availability constraints are as follows: The topological constraint is: for task chains Each pair of adjacent nodes in At any moment There must exist a corresponding directed edge. ; The task execution constraint is within the task chain. In the middle, node Actions only have nodes The task chain can only begin after the action is completed and the instruction or information is successfully transmitted; the execution of the task chain must follow strict task execution constraints. The functional coverage constraint is: task chain It must contain the minimum set of functional node types required to perform a specific task; for task chains Task chain The set of nodes in is , ; When the task chain For combat missions, All nodes in the set must have a node function type that is part of the strike mission type set. Elements in; ,in For reconnaissance nodes, As a decision-making node, For execution nodes, For the target node; When the task chain For reconnaissance missions, All nodes in the set must have a node function type that is a reconnaissance mission type. Elements in; ; When the task chain To ensure the success of the mission, All nodes in the set must have a node function type that corresponds to the guaranteed task type set. Elements in; ; To ensure the nodes are secure.

[0018] The successful transmission is defined as, for the node and It must be done at any time Simultaneously satisfying spatiotemporal constraints, transmission constraints, and time window constraints; The transmission constraint is: ; in, For nodes Startup time For nodes Startup time For nodes Processing latency; The time window constraint is: node and nodes The start time of transmission between Must fall on the node Time window and node Within the intersection of the time windows; ; in, For nodes Start time; For nodes End time; For nodes Start time; For nodes End time; If any node is in a fault state, maintenance state, or non-working time window, the transmission is considered to have failed.

[0019] The beneficial effects of this invention are as follows: By embedding spatiotemporal dynamic constraints as a core element into the node dynamic importance assessment logic, this invention quantifies the support contribution of nodes to the effective task chain under different environmental conditions by calculating dynamic data such as node position, running time window, and link delay in real time, combined with the constraints of the task chain. This dynamic assessment mechanism can accurately capture the dynamic importance changes under scenarios such as node maneuvering, time window shifting, and multi-source impact evolution, solving the problem of static indicators being disconnected from the dynamic task environment. This ensures that maintenance decisions always align with the task requirements of the equipment system, while breaking through the limitations of traditional static importance indicators. This invention uses limited support resources (maintenance units, repair time) to always allocate them to the nodes that contribute the most to the recovery of the equipment system's task capability. When the equipment system faces multiple node failures, by calculating the dynamic importance of each failed node, nodes with high dynamic importance are repaired first, which can quickly reconstruct critical functional paths and avoid the resource waste caused by the traditional "repair in fault order" approach. Simulation verification shows that this strategy can increase the recovery speed of the number of effective task chains in resource-constrained scenarios by more than 30%, significantly improving the utilization efficiency of support resources. The optimization method of this invention can effectively improve the mission reliability of equipment systems in multi-source confrontation scenarios. In random failure scenarios, by dynamically identifying the dynamic importance of randomly failed nodes, priority can be given to repairing the nodes most critical to mission chain coverage. In deliberate attack scenarios, when core nodes are targeted, high-dynamic-importance nodes can be quickly repaired to reconstruct the core framework of the system. In regional attack scenarios, by combining the functional complementarity and dynamic importance of nodes within a region, priority can be given to repairing the nodes within the region that provide the strongest support for cross-regional mission chains. This invention uses the number of effective mission chains as the core evaluation benchmark and directly links the dynamic importance of nodes to mission chain availability. The calculation logic of the dynamic importance of nodes is "the ratio of the number of effective mission chains when the node is normal / failed". This definition method fundamentally ensures the strong coupling between the dynamic importance index and mission execution requirements. Compared with traditional dynamic importance assessment based on topology, the index of this invention can better reflect the actual impact of nodes on mission completion capabilities, making the goal of maintenance decisions more focused on restoring mission execution capabilities rather than repairing the nodes themselves, and the evaluation logic is more in line with the mission value orientation of the equipment system. Attached Figure Description

[0020] Figure 1 A flowchart of a mission reliability optimization method for equipment systems based on importance, provided by this invention; Figure 2 The present invention provides a flowchart for optimizing the mission reliability of the equipment system. Figure 3 The invention provides a trend diagram of the change in the reliability of the equipment system mission under three attack modes before and after importance-based optimization over time; wherein... Figure 3 (a) The trend of target task reliability of a directed graph G(t) under random attack as a function of simulation time before and after the optimization method based on importance; Figure 3 (b) The trend of target task reliability of directed graph G(t) under deliberate attack as a function of simulation time before and after the optimization method based on importance; Figure 3 (c) The trend of target mission reliability of directed graph G(t) under region attack changes with simulation time before and after the optimization method based on importance; Figure 4 The present invention provides a comparison chart of the average target status of the equipment system before and after importance optimization under three attack modes; wherein, Figure 4 (a) The change of the target average state of the directed graph G(t) under random attack with simulation time before the importance-based optimization method; Figure 4 (b) shows the change of the target average state of the directed graph G(t) under random attack with simulation time after the importance-based optimization method. Figure 4 (c) The change of the target average state of the directed graph G(t) under intentional attack with simulation time based on the importance optimization method; Figure 4 (d) shows the change of the target average state of the directed graph G(t) under deliberate attack as a function of simulation time after the importance-based optimization method. Figure 4 (e) shows the change of the target average state of the directed graph G(t) under region attack as a function of simulation time before the importance-based optimization method. Figure 4 (f) shows the change of the target average state of the directed graph G(t) under region attack as a function of simulation time after the importance-based optimization method. Figure 5 The cumulative task chain count of the equipment system under the three attack modes provided by this invention is compared before and after importance optimization; among them, Figure 5 (a) The change in the number of task chains in the directed graph G(t) under random attack based on importance optimization method with simulation time; Figure 5 (b) The change in the number of task chains in the directed graph G(t) under deliberate attack based on importance optimization method with simulation time; Figure 5(c) The change in the number of task chains in the directed graph G(t) under region attack based on the importance optimization method with simulation time; Figure 6 The following is a comparison chart showing the completion rate of high-value target mission chains in the equipment system under three attack modes provided by this invention, based on importance optimization before and after optimization; among them, Figure 6 (a) The change in the completion rate of high-value targets in a directed graph G(t) under random attack based on importance optimization method before and after simulation time; Figure 6 (b) The change of the completion rate of high-value targets in a directed graph G(t) under deliberate attack based on importance optimization method before and after simulation time; Figure 6 (c) The change in the completion rate of high-value targets in the directed graph G(t) under region attack based on the importance optimization method before and after simulation time; Figure 7 The following is a comparison chart of the overall performance of the equipment system under three attack modes provided by this invention, based on importance optimization; wherein... Figure 7 (a) Comparison of the overall capabilities of a directed graph G(t) under random attacks based on importance optimization methods before and after; Figure 7 (b) Comparison of importance-based optimization methods for the comprehensive capabilities of a directed graph G(t) under deliberate attack; Figure 7 (c) Comparison of the overall capabilities of the directed graph G(t) under region attack based on the importance optimization method before and after. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0022] Currently, dynamic collaborative optimization of complex systems has become a research hotspot. Existing technologies include a digital twin-based intelligent collaborative management system for mobile machinery operations. This type of system is mainly applied in civilian logistics scenarios such as ports and yards. By constructing a spatiotemporal structure layer and a motion behavior layer, it utilizes "path conflict cones" and "congestion potential fields" to solve the path conflict and flow congestion problems of homogeneous mobile machinery (trucks, forklifts, etc.) within the physical operating space. Its core logic is based on cost function optimization of physical displacement, aiming to improve the overall operational throughput of the system. However, directly applying the above-mentioned civilian mobile machinery scheduling logic to heterogeneous equipment systems presents the following fundamental technical bottlenecks and incompatibilities: First, the node attributes shift from "homogeneous transport" to "heterogeneous emergence": Node attributes in mobile machinery systems are relatively simple and functionally equivalent, with their optimization core being obstacle avoidance at physical locations. In contrast, equipment systems are composed of highly heterogeneous nodes deeply coupled together, performing functions such as reconnaissance, decision-making, attack, and support; system efficiency relies on the logical closed loops of different functional nodes (such as OODA links), and the failure of a single node may lead to the collapse of the entire functional chain, rather than simple path blockage. Secondly, there's a shift in mission constraints from "physical paths" to "logical functional chains": Existing mission chains are essentially physical displacement paths based on yard topology. However, in high-intensity battlefield environments, the mission reliability of an equipment system depends on the integrity of its logical functional chains. Interactions between nodes involve not only positional movement but also real-time guidance of command and fire flows. Existing technologies lack measurement dimensions for "heterogeneous functional completion" and "logical chain reconstruction." Finally, the environmental characteristics are shifting from "business fluctuations" to "high-intensity combat damage": existing systems deal with random fluctuations in workload, aiming to "improve efficiency." However, equipment systems constantly face deliberate attacks from multiple sources, including firepower and electronic interference, leading to permanent physical damage to nodes or the instantaneous closure of mission windows. Traditional congestion perception logic cannot handle the drastic changes in system topology caused by battle damage, nor can it achieve reliable mission recovery after damage.

[0023] Addressing the disconnect between static importance metrics and the dynamic spatiotemporal environment, and the lack of adaptive scheduling algorithms under resource constraints, traditional equipment system maintenance optimization strategies present an importance-based method for optimizing equipment system task reliability. For example... Figure 1 As shown, it includes the following steps: Step 1: Establish a dynamic node importance assessment model based on task chains: Using the contribution of failed nodes to the mission reliability of the equipment system as a benchmark, quantitatively assess the dynamic importance index of each failed node. ; Step 1-1: Construct the equipment system as a dynamic directed graph with attributes. ; Directed graph for: ; in, A set of nodes; Let it be the set of edges; The set of attributes; the set of nodes includes several nodes; the nodes are the basic functional units constituting the equipment system; the edges The types include reconnaissance border, decision-making border, execution border, and support border; The attribute set A collection of all node attributes; node attributes for: ; in, For node identifiers; The node function types include reconnaissance node S, decision-making node D, execution node W, support node C, and target node T. For the node at time The instantaneous state of the node at time [time]. The instantaneous state includes the running state. Processing state Non-working status Fault state and maintenance status ; For the node at time The position vector; For the node at time The velocity vector; These are node capability parameters; node capability parameters include communication distance. and processing capacity ; Time window for nodes ; This refers to the node startup time; This is the end time of the node; A time window limits the time range within which a node can execute tasks; time. The position vector and velocity vector reflect the mobility of the node; The edge Represents a directed graph In the middle, when the node To the node Directed functional interaction paths between entities that satisfy spatiotemporal constraints; ; The spatiotemporal constraints are: ; That is, nodes and The Euclidean distance between them satisfies both the distance communication determination condition and the delay determination condition: The distance communication determination condition is the node. and The communication distance between them is less than the maximum communication distance. ; The delay determination condition is that within the maximum communication distance, the node... and Transmission delay between them It must be below the maximum threshold allowed by the task. And the interaction must occur on the node. and Common runtime window between Inside; The reconnaissance edge, decision-making edge, execution edge, and support edge are respectively: For the edge If node The type is target node T, node If the type is scout node S, then the edge For border reconnaissance; For the edge If node The type is decision node D, node If the edge is a node of any type other than the target node; then... For decision-making; For the edge If node The type is execution node W, node If the type is target node T, then the edge To execute the edge; For the edge If node The type is guaranteed node C, node For nodes of other types besides the target node, then the edge To protect the border; Directed graph The construction method is as follows: Based on mission requirements, deploy the basic functional units of a heterogeneous equipment system in space; using all basic functional units as nodes; traverse all nodes. compute nodes Position relative to all other nodes And Euclidean distance; if node If the Euclidean distances between the node and all other nodes simultaneously satisfy both the distance communication condition and the delay condition, then the node is activated. Edges between corresponding nodes; store activated edges in the edge set. If node If the Euclidean distance between a node and all other nodes does not satisfy the distance and delay conditions, then the node... There is no edge between the corresponding node and the directed graph; If, after any node is moved, the edges between the moved nodes satisfy the spatiotemporal constraints, then the corresponding edges between the nodes are activated, and the activated edges are added to the edge set. If the edges between nodes after the move do not satisfy the spatiotemporal constraints, then the edges will be removed from the edge set. Remove edges that do not satisfy the spatiotemporal constraints, thereby continuously updating the directed graph; Steps 1-2: Using a directed graph Build a task chain based on this; A mission chain is defined as a directed functional path within an equipment system, constructed to achieve specific mission objectives, such as reconnaissance of a specific area, engagement of a designated target, or support for a designated platform. It consists of a series of heterogeneous functional nodes interconnected by interactive edges, constructed and ordered according to specific mission logic and time requirements. In essence, a mission chain is a directed functional path within an equipment system, formed by a series of functional nodes connected through interactive links such as information flow, command flow, firepower flow, or resource flow, according to mission logic and timing requirements, in order to complete a specific reconnaissance / operation / support mission (such as reconnaissance of a specific area, engagement of a designated target, or support for a platform). That is, it starts from one or a group of nodes, passes through several functional nodes and corresponding edges (information flow, command flow, resource flow, firepower flow, etc.) in sequence, and finally reaches one or a group of nodes, forming a functional directed path.

[0024] The task chain is a directed graph. One or more directed paths for functions on the platform, subject to time constraints.

[0025] Task chain Directed graph The path above: by selecting Functional nodes in Dependency edges in, by Defined directions and constraints connect nodes into an ordered sequence to ultimately complete a specific task, and the entire process must satisfy the time and property constraints of the graph model; The task chain An ordered sequence of nodes used to perform specific tasks. ; in For task chains, For task chain The 1 node In the task chain Middle node The next node, For task chain Middle node and nodes The directed edges between them, Indicates from arrive Functional dependencies or execution connections, This represents the number of nodes in the task chain. Task chain The topology constraints, functional coverage constraints, task execution constraints, and availability constraints must be satisfied. The topological constraints are: For task chains Each pair of adjacent nodes in At any moment There must exist a corresponding directed edge. ; The functional coverage constraint is: task chain It must contain the minimum set of functional node types required to perform a specific task; for task chains Task chain The set of nodes in is , ; When the task chain For combat missions, All nodes in the set must have a node function type that is part of the strike mission type set. Elements in; ,in For reconnaissance nodes, As a decision-making node, For execution nodes, For the target node; When the task chain For reconnaissance missions, All nodes in the set must have a node function type that is a reconnaissance mission type. Elements in; ; When the task chain To ensure the success of the mission, All nodes in the set must have a node function type that corresponds to the guaranteed task type set. Elements in; ; To ensure the nodes; The task execution constraint is within the task chain. In the middle, node Actions only have nodes The task chain can only begin after the action is completed and the relevant information or instructions are successfully transmitted; the execution of the task chain must follow strict task execution constraints.

[0026] In the execution of the equipment system's mission chain, successful transmission refers to the node After internal processing is completed, the resulting information, instructions, firepower, or resource flows can be transmitted through directed edges. Complete and timely delivery to nodes and trigger the node The status of subsequent actions; side The interaction on the device is judged as a "successful transmission" and must be completed at a specific time. Simultaneously satisfying spatiotemporal constraints, transmission constraints, and time window constraints; The spatiotemporal constraints are nodes. and Satisfy spatiotemporal constraints; The transmission constraint is: ; in, For nodes Startup time For nodes Startup time For nodes Processing latency; The time window constraint is: node and nodes The start time of transmission between Must fall on the node Time window and node Within the intersection of the time windows; ; in, For nodes Start time; For nodes End time; For nodes Start time; For nodes End time; If any node is in a fault state, maintenance state, or non-working time window, the transmission is considered to have failed. Successful transmission requires not only that the source and destination nodes be within effective physical coverage, but also that the latency generated by the interaction meets the dynamic timing constraints of the task chain; and only if the edge has an indicator function. At this point, it is determined that the task information for this stage has been successfully transmitted, thus supporting the activation of subsequent task nodes. During task execution, all nodes and edges involved in the task chain must remain online and functioning normally within their respective time windows.

[0027] Steps 1-3: Calculate the current number of task chains based on the task chain: Define the sequence of functional nodes required by the task. Transform the "topology, timing, space, and availability" of the task chain definition into matrix filtering criteria. Reflect the flow of tasks in the equipment system network through matrix cascading. The input is the real-time status of the nodes. ,Location Task logic sequence Filter by node Distance and Delay Determination Adjacency matrix generation Recursively multiply. The final output is the total number of task chains. .

[0028] Step 1-3-1: Serialization of the target task logic flow: Based on the target mission (such as reconnaissance, strike, or support), the target mission is broken down into an ordered logical flow sequence; The ordered logical flow sequence for: ; in For the first The logical flow of each stage, The number of logic flows in an ordered logic flow sequence; ordered logic flow sequence The logical flow of each stage in the process is one of the node functional types; For the reason An ordered logical flow sequence consisting of functional stages The logic flow of a strike mission can be represented as "target detection S". Instruction Decision D Firepower Strike W”; the reconnaissance mission logic flow can be represented as “Target Detection S”. Instruction Decision D Target detection S”; the task logic flow can be represented as “detection node S / decision node D / execution node W”. Instruction Decision D Node C Detection node S / Decision node D / Execution node W”; Step 1-3-2: Based on the ordered logical flow sequence The Middle Each stage of logical flow, from the node set Extraction stage node subset stage node subset For a set of nodes The functional types of the middle node and the first All nodes with the same stage logic flow, i.e., exist; ; in, For nodes, For a set of nodes, Functions of type node functionality. For the first Each stage of logical flow Indicates the node at time [time]. It is alive and within the time window; Step 1-3-3: Constructing the adjacency matrix ; Adjacency matrix during construction phase The steps are as follows: Constructing a link matrix ,matrix The elements in are ,if Represents a node and There are logical or physical links between them; Construct a screening matrix , For filtering matrix Element; Formula for determining whether interactions between nodes satisfy spatiotemporal constraints:

[0029] in For filtering matrix The Line number Column elements, For indicator functions, if ,but ,like ,but , For Euclidean distance, This represents the maximum communication distance between nodes. For node processing latency, The time required for a link to complete its function (transmission delay). This represents the maximum allowable delay for this stage.

[0030] The first [item] is obtained through the Hadamard product. The stage adjacency matrix for each stage: ; Steps 1-3-4: Calculate the effective adjacency matrix for all stages. Based on all stage adjacency matrices Calculate the complete path count matrix for:

[0031] Where the matrix elements in Represents starting from the node To the destination node The number of all feasible paths that satisfy the task chain definition; The task chain full-path recursive calculation utilizes the path-passing property of matrix multiplication to synthesize the mappings of each stage in a chain, obtaining a complete path counting matrix from the initial functional stage to the final functional stage. ; Steps 1-3-5: Based on the complete path counting matrix Calculate the total number of task chains ; Total number of task chains The sum of all elements in the path counting matrix:

[0032] in It is a column vector of all 1s. It is a row vector consisting entirely of 1s; Steps 1-4: Simulate the nodes respectively This represents the task reliability of a directed graph in both running and fault states. Nodes At any moment Dynamic importance Defined as the ratio of the number of task chains in a directed graph where nodes are in the running state to the number of nodes in the failed state, the calculation formula is as follows: ; in, For nodes At any moment The dynamic importance, For nodes The reliability of the target task in a directed graph when it is in the running state. For nodes When in a fault state, the reliability of the target task in the directed graph. This represents the total number of task chains when the directed graph is created. The number of failed task chains. This represents the number of remaining task chains. Represents a node This represents the number of task chains during runtime. Represents a node The number of task chains in a fault state; In step one, the dynamic importance of all failed nodes is determined. After quantitative evaluation, a dynamic priority list of failure nodes is output. This list, serving as input to the reliability recovery model in step two, directly determines the maintenance scheduling decision. Resource orientation.

[0033] Step 2: Construct a reliability recovery model under the constraints of limited maintenance resources: The maintenance and support process of the equipment system is abstracted into a dynamic constraint optimization problem, and the reliability recovery target is quantified by establishing a mathematical model.

[0034] Step two introduces the dynamic importance of the failed nodes calculated in step one. This solves the "repair sequence optimization problem" under the constraint of limited guarantee resources. Specifically, the reliability recovery objective function in step two... In fact, it involves maintaining scheduling decisions. Prioritize changing dynamic importance The highest failure node state (from the failure state) Restore to running state To maximize the mission chain within the equipment system. Supply, thereby improving the overall reliability of the mission. The technical objective is to ensure that the recovery trajectory of the equipment system's mission completion capability remains on the optimal path after being subjected to random or deliberate shocks. This evaluation-driven optimization logic ensures that the system's mission completion capability recovery trajectory remains on the optimal path.

[0035] Step 2-1: Quantitative characterization of task reliability; Building Task Reliability This represents the probability that the equipment system will successfully complete its designated mission objective; its quantitative calculation is based on Monte Carlo simulation statistics, and the calculation formula is as follows: ; In the formula, Total number of Monte Carlo simulations; For the first The function indicating the success of the task in this simulation run; where... The criterion for determination is time. Total number of task chains in a directed graph ; Step 2-2: Construct the reliability recovery objective function; With the objective of maximizing the cumulative mission reliability across the entire mission profile, the following reliability recovery objective function formula is established: ; in, The total known task duration; For mission reliability, It is the set of instantaneous states of all nodes at all times; To maintain scheduling decisions, this function aims to dynamically repair nodes with high dynamic importance, enabling the system to quickly recover its task execution capability after being impacted; the reliability recovery objective function is a single-objective optimization function.

[0036] Steps 2-3: Determine the maintenance scheduling decision variables; Maintenance scheduling decision To ensure that each available guarantee node C in the guarantee resource pool is at time [time missing] Determine which failed node to be assigned for repair, where the guaranteed resource pool consists of all nodes of the guaranteed node type. Steps 2-4: Maintaining scheduling decisions The core maintenance constraint set; Maintaining resource constraints: at any time The number of nodes in maintenance mode at the same time cannot exceed the maximum number of maintenance tasks that can be performed. .

[0037] ; Maintenance time constraint: any node Each repair requires a non-zero repair time. If the node At the point of time The repair process begins, and the time it takes for it to return to running state is... satisfy: ; Step 3: Execute the dynamic repair scheduling algorithm based on dynamic importance priority: Based on the real-time updated dynamic importance ranking of nodes, dynamically allocate maintenance resources and update node status to achieve the fastest recovery of task reliability.

[0038] Step 3-1: Real-time status monitoring and candidate repair node screening; Real-time traversal of the node state set of a directed graph It identifies the current state as "faulty". "Status and not in "maintenance status" "The set of nodes is marked as a candidate state to be repaired; the set of all nodes marked as candidate states to be repaired is the candidate node set." ; Step 3-2: Dynamic quantification and priority ranking of candidate nodes; The evaluation model established in step one is invoked to calculate the candidate node set in real time. The current dynamic importance of each node. In the candidate node set In the process, nodes are sorted from highest to lowest dynamic importance to obtain a dynamic priority list. ; Step 3-3: Competition for maintenance resources and optimal allocation decision; Search the resource pool to determine the number of currently available repair units. Is it greater than zero, where .like If the candidate list is not empty, then the dynamic priority list will be used. The repair units are assigned to the candidate node set in the following order. The middle node; the resource pool consists of all nodes whose node type is "Guaranteed Node". Steps 3-4: Closed-loop update of node status and performance recovery; Step 3-4-1, State Transition: Once the candidate node set is complete... The middle node is assigned to the backup node in the backup resource pool for maintenance, and a candidate node set is provided. The state of the intermediate node is changed from the fault state. Immediately switch to maintenance mode ; Step 3-4-2, Delay Simulation: Record the set of candidate nodes for ensuring node repair. After the repair time of the intermediate nodes meets the maintenance time constraint, the candidate node set is determined. The middle node status is updated to running state. ; Step 3-4-3, Capability Reorganization: The Repaired Set of Candidate Nodes The intermediate node reconnects to the network, participates in the construction of a new task chain, and uses an objective optimization function to improve task reliability. This maximizes the equipment system's mission completion capability, enabling dynamic injection of mission capabilities.

[0039] The equipment system mission reliability optimization method proposed in this invention first identifies the set of failed nodes caused by external shocks (random, deliberate, or area attacks) by real-time monitoring of the node status of the equipment system under dynamic combat environment. Then, it calls the evaluation model described in step one and, based on the topology, function, timing, and availability constraints defined in the mission chain, uses a matrix recursive algorithm to quantitatively calculate the marginal contribution of each failed node to the system's mission capability supply, i.e., the dynamic importance index. Next, this dynamic importance is used as the core decision-making basis and input into the resource-constrained recovery model described in step two. The scheduling algorithm described in step three then outputs the optimal maintenance plan under the dual constraints of ensuring resource limits and repair time. Finally, by prioritizing the repair of nodes with high dynamic importance, the number of mission chains in the equipment system is increased. A rapid recovery, thereby maximizing cumulative task reliability. The optimization objective.

[0040] To verify the effectiveness of the above logic under actual mission profiles, this embodiment employs a combination of discrete-time step simulation and Monte Carlo simulation. The flowchart of the importance-based equipment system mission reliability optimization method is as follows: Figure 2 As shown, the specific steps are as follows: Step 1: Input simulation parameters. Initial directed graph of the equipment system. Total number of simulations Total task duration Functional logic sequence of each stage Delay threshold Maintenance resource limit and node repair time wait; Step 2: Initialize the simulation data and set the simulation iteration count. Number of successful tasks ; Step 3: Determine the number of simulation iterations Has the termination number been reached? If the termination number is reached Then end the simulation and proceed to step 11; otherwise, let Proceed to step 4; Step 4: Initialize the simulation time. The total simulation time is ; Step 5: Multi-source Uncertainty Impact Simulation. The Monte Carlo method is applied to simulate attack modes on equipment system nodes, including random attacks, deliberate attacks, and area attacks, with real-time updates of the position coordinates of each node. and state set Identify the fault state Nodes; Step 6: Scheduling decision based on dynamic importance.

[0041] Step 61 (corresponding to Step 1): Identify the set of nodes to be repaired, i.e., the failed nodes, and call the dynamic importance assessment model to calculate the dynamic importance of the failed nodes. ; Step 62 (corresponding to Step 3): Based on the dynamic importance of the failure node The ranking system establishes a dynamic priority list of failed nodes. Under the premise of meeting resource constraints, it allocates maintenance resources and updates the status of failed nodes to "maintain". ; Step 7: Calculate the total number of task chains that satisfy all topological, functional, and time constraints using matrix recursion operations. The calculation formula is: ; Step 8: Determine the current time. Has the total task duration been reached? :like Then let Update the status of the repaired node and proceed to step 5; otherwise, proceed to step 9. Step 9: Determine if the time condition is met. Total number of mission chains within the equipment system If the condition is met, proceed to step 10; otherwise, proceed to step 11. Step 10: Update the number of successful tasks. If step 9 determines that the task is successful, then... Otherwise, keep constant; Step 11: Determine if the number of iterations has reached the termination number: If it has, end the iteration and proceed to step 12; otherwise, proceed to step 3. Step 12: Calculate the equipment system's performance based on the full sample statistical results. Real-time task reliability:

[0042] In the formula, For the equipment system at all times Task reliability, The number of times the task was successfully completed. This represents the total number of simulations. This number is used as the final metric for evaluating the optimization method of this invention.

[0043] As a further limitation of the present invention, the attack scenario is selected from at least one of the following: Random attack: Simulates random impact, where each node in the system has the same probability of failure; Intentional attack: The highest node in the targeted strike equipment system; Area attack: Simulates spatial impact, causing nodes and edges within a specific area to become completely ineffective.

[0044] This invention provides an analysis and study of scenarios under three attack modes. Please refer to [link / reference]. Figure 3 This graph shows the trend of equipment system mission reliability changing over time before and after importance-based optimization under three different attack modes. The Y-axis represents "mission reliability". In the Monte Carlo simulation, the X-axis represents the proportion of simulations that still meet the preset mission success criteria at time t. The X-axis represents the simulation time (in hours). In the random attack scenario, the blue curve before optimization shows that under continuous random failures, the system's mission reliability begins to decline after 0.25 hours and completely drops to zero after 0.5 hours, ultimately leading to the system's "complete exhaustion." The purple curve after optimization demonstrates significant resilience. The time point at which performance begins to decline is delayed from 0.25 hours to 0.3 hours, preventing complete system collapse and stabilizing mission reliability at approximately 40%, forming a dynamic equilibrium for sustainable operation. In the deliberate attack scenario, before optimization, under precise strikes targeting key nodes, the system's reliability begins to decline sharply from 0.1 hours and completely collapses to zero within 0.3 hours, exhibiting high vulnerability. After optimization, prioritizing the repair of attacked core nodes significantly slowed the rate of reliability decline, maintaining the final reliability level at approximately 16%, successfully preventing complete system paralysis. In a regional attack scenario, before optimization, the system's reliability instantly dropped to zero after approximately 0.1 hours of devastating regional damage, and it could never recover. After optimization, the system's overall capability declined slowly through a process of gradual recovery and subsequent damage, but the rate of decline was much slower than before optimization. Reliability only approached zero after approximately 0.45 hours, and its effective combat time was extended by more than double (by 125%).

[0045] Please see Figure 4 This chart compares the evolution of 40 target states under three different attack modes. The Y-axis represents the average number of targets in different states, and the X-axis represents the simulation time (hours). The different colored bars represent the different target states listed from top to bottom: leaked, destroyed, engaged, assigned to weapons, tracked, detected, and active. The black dashed line represents the cumulative total number of targets entering the battlefield. By comparing the before-and-after comparison charts under various attack modes, the effectiveness of the system was significantly improved. Under random attacks, the optimized system destroyed approximately 29 targets, an increase of nearly 93% compared to the 15 targets before optimization. Under deliberate attacks, the number of targets destroyed increased from 10 to 22, an improvement of 120%. The importance-based optimization method reduced the number of targets that would have broken through defenses from 30 to 18, successfully intercepting 12 targets that would have broken through defenses, thus increasing the mission success rate from 25% to 55%. Under area attacks, the optimized system was able to continue fighting after the attack, destroying an additional 3 targets, increasing the battle result by 75%. This indicates that the optimization method can effectively maintain and restore the system's mission chain, effectively cope with continuous internal node failures, and ensure the success of critical defense missions.

[0046] Please see Figure 5Under three typical attack scenarios, a detailed comparative analysis of the average cumulative task chain evolution process before and after equipment system optimization is conducted. In each graph, the dashed line represents the number of task chains "formed" (detection, tracking, and weapon allocation completed), and the solid line represents the number of "completed" (target successfully destroyed). The pink / yellow curve represents the system performance "before optimization" (without repair), while the purple / blue curve represents the system performance "after optimization" after the addition of an importance-based intelligent repair strategy. Under random attack scenarios, the system before optimization completed approximately 15 task chains within its lifecycle, while the system after optimization completed approximately 30 task chains. The optimization method increased the system's task completion rate by approximately 100%. Furthermore, the system before optimization essentially lost its mission capability after 0.6 hours, while the system after optimization maintained a highly efficient task chain generation capability throughout the simulation period of up to 1 hour, indicating that the optimization method significantly extended the system's effective mission time under harsh environments. Under deliberate attack scenarios, the system before optimization collapsed much faster than in random failure scenarios when facing attacks specifically targeting key nodes. After approximately 0.3 hours, the curve had essentially flattened out, indicating that the core network had been compromised. Only about 10 task chains were ultimately completed. The optimized repair system prioritized repairing the attacked critical nodes. Although the curve's growth slope was lower than in the random attack scenario, it still maintained continuous growth. Approximately 22 task chains were ultimately completed, with the optimization method increasing the system's task completion rate by about 120%. In the regional attack scenario, the system's task capability experienced a precipitous drop before optimization. After the attack, the curve immediately flattened out, and only about 4 task chains were ultimately completed, all of which were completed before the attack. The optimized repair system assessed the importance of all failed nodes within the attack area and repaired them, ultimately completing 7 task chains. The optimization method increased the system's task completion rate by about 75%.

[0047] Please see Figure 6Under three attack scenarios, a comprehensive comparative analysis was conducted on the completion rates of high-value targets before and after equipment system optimization. The purple diamond curve represents the system performance "before optimization" (without repair), while the blue triangle curve represents the system performance "after optimization" after incorporating an importance-based optimization method. In the random attack scenario, the purple curve before optimization showed a clear and continuous downward trend after initial fluctuations, eventually stabilizing at a high-value target task chain completion rate of approximately 51%. The blue curve after optimization, after experiencing initial fluctuations, successfully suppressed the downward trend and stabilized and rebounded, maintaining the health of the equipment system. Ultimately, the completion rate stabilized at a relatively high level of approximately 81%, with the optimization method improving the core task completion rate by 30%. In the deliberate attack scenario, before optimization, facing attacks specifically targeting key nodes, the purple curve decreased much faster and more deeply than in the random attack scenario, eventually stabilizing at an extremely low level of only about 35%, essentially paralyzing the core functions of the system. The optimized blue curve demonstrated strong resilience, ultimately maintaining a completion rate of approximately 61%. The optimization method improved the core task completion rate by 26%, a relative increase of over 74%, demonstrating significant effectiveness. In the area attack scenario, the purple curve experienced a precipitous drop, declining to only about 8%, indicating the system almost completely lost its ability to complete tasks after the attack. The optimized blue curve successfully mitigated the damage and slightly recovered, ultimately stabilizing the completion rate at approximately 15%. The optimization method improved the core task completion rate by 7%, achieving nearly double the performance recovery.

[0048] To comprehensively evaluate the importance-based optimization method proposed in this paper, this section constructs a comprehensive performance evaluation framework that includes five core indicators: mission reliability, high-value target destruction rate, rapid interception capability, number of mission chains, and mission reliability gain. Please refer to [link / reference]. Figure 7 The optimization effects are compared under three typical attack scenarios. The dark solid-line polygon represents the overall system performance before optimization, while the light-filled polygon represents the overall system performance after optimization. The larger the area covered by the polygon, the better the overall system performance. The comparison of different indicators before and after optimization is shown in Table 1.

[0049] Table 1: Comparison of different indicators before and after optimization

[0050] Random Attack Scenario: Before optimization, the system's performance indicators significantly declined after a random attack. The final task reliability was only 0, the high-value target destruction rate was only 50%, and the number of task chains was 10. After optimization, the final task reliability improved to 0.4 (a 40% increase), the high-value target destruction rate reached 80%, and the number of task chains increased to 22 (a 107% increase). The area of ​​the polygon after optimization was significantly larger than before, demonstrating the effectiveness of the optimization method in dealing with random failures. Intentional Attack Scenario: Before optimization, the system's performance severely collapsed under precise attacks targeting critical nodes. The final task reliability was 0, the number of task chains was 10, and the high-value target completion rate was only 33%. After optimization, the optimization method effectively countered intentional attacks. By prioritizing the repair of critical nodes, the optimized system maintained task reliability at 0.17 (a 17% increase), increased the number of task chains to 22 (a 120% increase), and also increased the high-value target destruction rate to 60%. Area Attack Scenario: Before optimization, the system's performance instantly dropped to zero after suffering a devastating area attack. The radar chart shows that the unoptimized light-colored polygon had shrunk to almost zero, with all indicators at their lowest levels. After optimization, although the final mission reliability remained at 0 after suffering severe damage, the optimized system achieved a 78.5% increase in mission reliability by repairing key nodes, the number of mission chains recovered to 7, and the high-value target destruction rate recovered to 15%. Although the absolute values ​​are not high, the optimized dark-colored polygon still clearly envelops the unoptimized polygon, indicating that the optimization method remains an effective means of improving the mission reliability of the indicator system even in extremely severe environments.

[0051] The above analysis results systematically verify the effectiveness and applicability of the proposed optimization method from multiple dimensions. This method not only significantly improves the mission reliability, number of mission chains, and completion rate of high-value targets of the equipment system under different types of perturbations, but also reveals that after incorporating the optimization method, the equipment system can be transformed from a fragile and easily collapsing structure into an equipment system with damage tolerance and recovery capabilities. This has important theoretical and practical guiding significance for the design and application of equipment systems in future high-confrontation environments.

Claims

1. A method for optimizing the mission reliability of an equipment system based on importance, characterized in that, Includes the following steps: Step 1: Construct a directed graph based on the basic functional units of the equipment system. With directed graph Based on this, a task chain is constructed, breaking down the target task into an ordered logical flow sequence; the ordered logical flow sequence includes... The logical flow of each stage; simulating and calculating the directed graph based on the ordered logical flow sequence and the node set in the directed graph. Middle node At any moment Dynamic importance ; Step 2: Construct a reliability recovery model; the reliability recovery model includes task reliability. and the reliability recovery objective function; The reliability of the construction task for: ； In the formula, Total number of Monte Carlo simulations; For the first Next calculation time The result of the total number of task chains in the directed graph; if the total number of task chains is greater than zero, then... The value is 1; if the total number of task chains is less than or equal to zero, then The value of is 1; The reliability recovery objective function is: ； in, The total known task duration; It is the set of instantaneous states of all nodes at all times; To maintain scheduling decisions; the maintenance of scheduling decisions To ensure that each available guarantee node C in the guarantee resource pool is at time [time missing] The process involves determining which failed node to allocate resources for repair, where the resource pool comprises all nodes of type "protected node"; the maintenance scheduling decision... The core maintenance constraint set must be met; the core maintenance constraint set includes maintenance resource constraints and maintenance time constraints. To represent the task reliability function, The value is determined by time. , and maintenance scheduling decisions To be determined jointly; Step 3: Real-time traversal of the nodes of the directed graph at time... The instantaneous state is used to identify the current fault state. And it is not in maintenance condition. The identified nodes are marked as candidate states to be repaired; the set of all nodes marked as candidate states to be repaired is the candidate node set. Real-time calculation of candidate node set The current dynamic importance of each node. In the candidate node set In the process, nodes are sorted from highest to lowest dynamic importance to obtain a dynamic priority list. ; Search the resource pool for available maintenance units. And the set of candidate nodes If not empty, then follow the dynamic priority list. The repair units are assigned to the candidate node set in the following order. Middle node, number of currently available maintenance units or candidate node set If empty, no maintenance unit will be allocated; the resource pool consists of all nodes of the guaranteed node type; if the candidate node set... The middle node is being repaired by the guaranteed nodes in the resource pool, and a candidate node set is being developed. The state of the intermediate node is changed from the fault state. Immediately switch to maintenance mode Record the set of candidate nodes for node repair. After the repair time of the intermediate nodes meets the maintenance time constraint, the candidate node set is determined. The middle node status is updated to running state. The repaired set of candidate nodes The middle node is reconnected to the directed graph, and the directed graph is updated in real time. Using the reliability recovery objective function to assess task reliability Optimize to improve task reliability To maximize the mission completion capability of the equipment system.

2. The method for optimizing the mission reliability of an equipment system based on importance, as described in claim 1, is characterized in that... The construction of directed graphs The steps of task chain M are as follows: Step S1-1-1: Construct a directed graph based on the basic functional units of the equipment system. ; Based on mission requirements, deploy the basic functional units of a heterogeneous equipment system in space; using all basic functional units as nodes; traverse all nodes. compute nodes Position relative to all other nodes And Euclidean distance; if node If the Euclidean distances between the node and all other nodes simultaneously satisfy both the distance communication condition and the delay condition, then the node is activated. The edge between the node and the corresponding node; if the node If the Euclidean distance between a node and all other nodes does not satisfy the distance and delay conditions, then the node... There is no edge between the corresponding node and the edge; Construct a directed graph based on all nodes and active edges. for: ； in, A set of nodes; The set of all active edges; The set of attributes; the set of nodes includes all nodes; the set of edges For all edges The set; the edges The types include reconnaissance edges, decision-making edges, execution edges, and support edges; after construction, the directed graph is updated in real time. ; The attribute set A collection of all node attributes; node attributes for: ； in, For node identifiers; The node function types include reconnaissance node S, decision-making node D, execution node W, support node C, and target node T. For the node at time The instantaneous state of the node at time [time]. The instantaneous state includes the running state. Processing state Non-working status Fault state and maintenance status ; For the node at time The position vector; For the node at time The velocity vector; These are node capability parameters; node capability parameters include communication distance. and processing capacity ; Time window for nodes ; This refers to the node startup time; This is the end time of the node; The edge Represents a directed graph In the middle, when the node To the node Directed functional interaction paths that satisfy the distance communication determination condition and the delay determination condition; ; Step S1-1-2: Using a directed graph Build a task chain based on this; The task chain for: ； in For task chains, For task chain The 1 node In the task chain Middle node The next node, For task chain Middle node and nodes The directed edges between them, Indicates from arrive Functional dependencies or execution connections, The number of nodes in the task chain; the task chain It must also satisfy topological constraints, functional coverage constraints, task execution constraints, and availability constraints.

3. The method for optimizing the mission reliability of an equipment system based on importance, as described in claim 1, is characterized in that... The target task is decomposed into an ordered logical flow sequence; the ordered logical flow sequence includes... The logical flow of each stage; simulating and calculating the directed graph based on the ordered logical flow sequence and the node set in the directed graph. Middle node At any moment Dynamic importance The steps are as follows: Step S1-2-1: Based on the ordered logical flow sequence The Middle Each stage of logical flow, from the node set Extraction stage node subset stage node subset For a set of nodes The functional types of the middle node and the first All nodes with the same stage logic flow; Step S1-2-2: Based on the subset of stage nodes Constructing an adjacency matrix ; Adjacency Matrix Obtained through the Hadamard product, i.e.: ； in, For link matrix; For filtering matrix; For Hadamard product; Link Matrix The elements in are ,if Represents a node and There are logical or physical links between them; Filtering matrix elements in for: ； in For filtering matrix The Line 1 Column elements, For indicator functions, if ,but ,like ,but , For Euclidean distance, This represents the maximum communication distance between nodes. For node processing latency, The time required for the task chain to complete its function. For the first The maximum allowable delay for each stage; Step S1-2-3: Calculate the adjacency matrix for all stages. Based on the adjacency matrix of all stages Calculate the complete path count matrix for: ； The complete path count matrix elements in Represents starting from the node To the destination node The number of all feasible paths that satisfy the task chain definition; Step S1-2-4: Based on the complete path counting matrix Calculate the total number of task chains ; Total number of task chains The sum of all elements in the path counting matrix: in It is a column vector of all 1s. It is a row vector consisting entirely of 1s; Step S1-2-5: Simulate the operation of the computing nodes At any moment Dynamic importance Dynamic importance For nodes The ratio of the number of task chains in the running state to the number of task chains in the fault state in a directed graph is calculated using the following formula: ； in, For nodes At any moment The dynamic importance, For nodes The reliability of the target task in a directed graph when it is in the running state. For nodes When in a fault state, the reliability of the target task in the directed graph. The total number of task chains when the directed graph is created. The number of failed task chains. Represents a node This represents the number of task chains during runtime. Represents a node This represents the number of task chains in a faulty state.

4. The method for optimizing the mission reliability of an equipment system based on importance, as described in claim 1, is characterized in that... The real-time updated directed graph The steps are as follows: If a directed graph If, after any node is moved, the edges between the moved nodes simultaneously satisfy both the distance communication condition and the delay condition, then the corresponding edges between the nodes are activated, and the activated edges are added to the edge set. If the edges between nodes after the move do not satisfy the distance communication and delay judgment conditions, then the edges will be removed from the edge set. Remove edges that do not meet the distance communication determination condition or the delay determination condition; thus enabling the directed graph to be continuously updated.

5. The method for optimizing the mission reliability of an equipment system based on importance according to claim 1, characterized in that, The maintenance resource constraint is: at any time The number of nodes in maintenance mode at the same time cannot exceed the maximum number of maintenance tasks that can be performed. .

6. The method for optimizing the mission reliability of an equipment system based on importance according to claim 1, characterized in that, The maintenance time constraint is: if node At the point of time Start repair, then the node Time to restore running state satisfy: ； Maintenance time constraints are for nodes The repair time is a non-zero repair duration. .

7. The method for optimizing the mission reliability of an equipment system based on importance, as described in claim 2, is characterized in that... The node and The Euclidean distance between them satisfies both the distance communication determination condition and the delay determination condition as follows: The distance communication determination condition is the node. and The communication distance between them is less than the maximum communication distance. ; The delay determination condition is that within the maximum communication distance, the node... and Transmission delay between them It must be below the maximum threshold allowed by the task. And the interaction must occur on the node. and Common runtime window between Inside.

8. The method for optimizing the mission reliability of an equipment system based on importance according to claim 2, characterized in that, The topological constraints, functional coverage constraints, task execution constraints, and availability constraints are as follows: The topological constraint is: for task chains Each pair of adjacent nodes in At any moment There must exist a corresponding directed edge. ; The task execution constraint is within the task chain. In the middle, node Actions only have nodes The task chain can only begin after the action is completed and the instruction or information is successfully transmitted; the execution of the task chain must follow strict task execution constraints. The functional coverage constraint is: task chain It must contain the minimum set of functional node types required to perform a specific task; for task chains Task chain The set of nodes in is , ; When the task chain For combat missions, All nodes in the set must have a node function type that is part of the strike mission type set. Elements in; ,in For reconnaissance nodes, As a decision-making node, For execution nodes, For the target node; When the task chain For reconnaissance missions, All nodes in the set must have a node function type that is a reconnaissance mission type. Elements in; ; When the task chain To ensure the success of the mission, All nodes in the set must have a node function type that corresponds to the guaranteed task type set. Elements in; ; To ensure the nodes are secure.

9. The method for optimizing the mission reliability of an equipment system based on importance, as described in claim 8, is characterized in that... The successful transmission is for the node and It must be done at any time It simultaneously satisfies the distance communication determination condition, the delay determination condition, the transmission constraint, and the time window constraint; The transmission constraint is: ； in, For nodes Startup time For nodes Startup time For nodes Processing latency; The time window constraint is: node and nodes The start time of transmission between Must fall on the node Time window and node Within the intersection of the time windows; ； in, For nodes Start time; For nodes End time; For nodes Start time; For nodes End time; If any node is in a fault state, maintenance state, or non-working time window, the transmission is considered to have failed.

10. A terminal device, comprising a processor, a memory, and a computer program stored in the memory; characterized in that, When the processor executes the computer program, it implements the method for optimizing the reliability of equipment system tasks based on importance, as described in any one of claims 1-9.

Images