Unmanned aerial vehicle cluster weapon target collaborative allocation method and system under space-time constraint

Abstract

The invention provides an unmanned aerial vehicle cluster weapon target collaborative allocation method and system under space-time constraint, and the method comprises the steps: defining a collaborative allocation object, constructing a collaborative combat scene target function, defining a collaborative combat scene constraint condition, and finally solving the target function through employing a multi-gene population parallel ant colony algorithm, encoding the weapon set, the target set and the unmanned aerial vehicle set into a weapon gene sequence, a target gene sequence and an unmanned aerial vehicle gene sequence respectively to generate an initial population, and performing staged joint optimization by fusing a pheromone guiding mechanism of an ant colony algorithm and crossover mutation operation of a genetic algorithm, the optimal cooperative allocation scheme is searched when the constraint condition is satisfied, the unmanned aerial vehicle task allocation result can be better obtained and optimized through the cooperative allocation mode, and the unmanned aerial vehicle cooperative allocation efficiency is improved.

Description

Technical Field

[0001] This article belongs to the field of unmanned aerial vehicle (UAV) planning technology, specifically involving a method and system for collaborative allocation of weapon targets in UAV swarms under spatiotemporal constraints. Background Technology

[0002] Drone technology has developed rapidly in recent years and has been widely applied in various fields such as military, civilian, scientific research, and commerce. With advancements in drone technology and reductions in cost, drone swarm technology has gradually attracted widespread attention. Drone swarm technology refers to multiple drones working collaboratively to complete one or more specific tasks. In traditional drone operation, a single drone may be limited by various factors when completing a specific task, such as battery life, flight speed, and payload. Drone swarm technology, however, can fully leverage the advantages of a cluster by working collaboratively with multiple drones, thereby improving mission efficiency and success rates.

[0003] The planning of unmanned aerial vehicle (UAV) swarm operations is a crucial and highly relevant topic in the military field. With the increasing complexity and diversity of modern warfare, traditional combat methods are no longer sufficient to meet the demands for rapid acquisition and real-time monitoring of enemy situation, terrain, and battlefield conditions. UAVs, as a flexible, covert, and efficient combat platform, offer entirely new possibilities for military mission execution. First, UAVs possess excellent reconnaissance and surveillance capabilities, enabling them to perform various reconnaissance missions on the battlefield, acquiring vital information such as enemy situation, terrain, and weather conditions, providing reliable intelligence support for operational command and decision-making. Second, UAVs can execute precision strike missions while minimizing casualties among friendly forces and civilians while maintaining combat effectiveness.

[0004] Traditional UAV mission allocation methods primarily consist of two parts: target allocation and weapon allocation. Target allocation typically employs the Multiple Traveling Salesman Problem (MTSP) model, which requires consideration of constraints such as UAV range and time, and the rational planning of mission sequence. Weapon allocation mainly utilizes the Weapon-Target Assignment (WTA) problem model, which models the different weapon types and their varying damage probabilities to targets as a mathematical problem, and solves it using optimization algorithms. However, this approach fails to consider the constraints and synergies between the two, resulting in certain shortcomings. Summary of the Invention

[0005] To address the aforementioned problems in existing technologies, this paper aims to provide a method and system for collaborative allocation of weapon targets in UAV swarms under spatiotemporal constraints, which can improve the efficiency of UAV collaborative allocation.

[0006] To solve the above-mentioned technical problems, the specific technical solution presented in this paper is as follows: On the one hand, this paper provides a method for cooperative allocation of weapon targets in UAV swarms under spatiotemporal constraints, the method comprising: S10: Define a collaborative allocation object, which includes a set of drones, a set of targets, and a set of weapons; S20: Construct a cooperative combat scenario objective function, which is used to simultaneously optimize the following four objectives: target survival probability index, UAV total flight range index, mission completion time index, and UAV survival probability index. S30: Define the constraints for the collaborative combat scenario, which include at least time constraints, space constraints, and UAV capability constraints; S40: The objective function is solved using a multi-gene population parallel ant colony algorithm, wherein the weapon set, target set, and UAV set are encoded as weapon gene sequences, target gene sequences, and UAV gene sequences, respectively, to generate an initial population. The pheromone guidance mechanism of the ant colony algorithm and the crossover and mutation operation of the genetic algorithm are combined in stages for joint optimization. Under the constraints, the optimal cooperative allocation scheme among UAVs, weapons, and targets is searched. S50: If a drone failure is detected during mission execution, a dynamic reallocation strategy is triggered, which only performs local re-optimization on the affected target and the remaining drones to generate a new collaborative allocation scheme.

[0007] Furthermore, the objective function of the cooperative combat scenario is expressed by the following formula: ; in, This represents an indicator of the probability of the target surviving. This indicates the total range of the drone. This indicates the timeframe for task planning. This indicates the survival probability index of drones. These represent the weights of the target survival probability, the total flight range of the UAV, the mission planning time, and the UAV survival probability, respectively. ; , Indicates the probability of the target surviving. , Indicates the maximum range of the drone. This indicates that there is no total number of people. Indicates the total number of targets. Indicates the total number of weapons. Indicate weapon type For the target The probability of damage, Indicates assignment to target of The number of type weapons Indicate target j The weights; , Indicates the total flight range of the drone. ; Indicates the first A drone from the target To the target Euclidean distance, Represents a binary decision variable; , Indicates the task planning time. , Indicates the first A drone from the target fly to target Flight time, Indicate target The available time window Indicates the drone's cruising speed; , Indicates the survival probability of a drone. , Indicates the maximum range of the drone. Indicates that the drone was targeted The probability of damage.

[0008] Furthermore, the constraints of the cooperative combat scenario are expressed by the following formula: Time constraints: ; ; ; ; in, Indicates the first The drone reached the target. Time, Indicates the specified task start time; Indicates the first A drone from the target fly to target Flight time, Indicates the first A drone flies from the starting point to the target. i Flight time, Indicate target The available time window Represents a binary decision variable. Indicates the first i No more than one drone Complete the task. This indicates that the p-th drone starts from the target. fly to target Binary decision variables on the flight path, This indicates that the p-th drone flies from the starting point to the target. The binary decision variables on the flight path; Spatial constraints: ; ; in, and This means that the p-th drone takes off from the starting point and needs to fly back to the starting point after completing its mission; Unmanned aerial vehicle (UAV) capability constraints: ; ; in, Indicates the first The drone was assigned to the attack target The number of weapons, Indicates the maximum ammunition load of the drone. Indicates the maximum range of the drone. Indicates the first A drone from the target To the target Euclidean distance, This represents a binary decision variable.

[0009] Furthermore, the initial population in step S4 is obtained through the following steps: The weapon set, target set, and drone set are encoded into weapon gene sequences, target gene sequences, and drone gene sequences, respectively. Random deletion and replacement operations are performed on the target gene sequence, and random sorting is performed on the UAV gene sequence to generate an initial population.

[0010] Further, step S4 includes: Based on the aforementioned constraints, the UAV gene sequences are optimized to determine the mission access order for each UAV. The assignment relationship between fixed UAVs and targets is established, and the UAV gene sequence and the target gene sequence are combined into a multidimensional gene sequence. Based on the aforementioned multidimensional gene sequence, and in conjunction with the weapon parameters of each weapon, the weapon gene sequence is optimized; The optimized parallel gene sequences are fused together, and the candidate solutions are sorted using the objective function to perform Pareto non-dominated sorting. Elite solutions are then selected, and the pheromones are updated based on the elite solutions. Each generation of elite solutions is directly passed to the next generation of the population, and the population is updated through crossover and mutation operations. This process is iterated and optimized until the convergence condition is met.

[0011] Furthermore, the pheromone is updated in the following manner: ; ; in, This represents the pheromone along the execution path from target i to target j; Indicates the evaporation rate. Indicates the initial pheromone evaporation rate. This represents the final pheromone evaporation rate, and satisfies 0 < < <1; t Represents iterative algebra, >0 represents the decay rate parameter, used to control the transition speed of the evaporation rate from the exploration stage to the development stage.

[0012] Furthermore, the multi-gene population parallel ant colony algorithm performs Pareto non-dominated sorting on candidate solutions in each iteration, retains the non-dominated solutions as the elite solution set, and performs a pheromone enhancement operation based on the elite solution set, wherein the pheromone increment Δ It is inversely proportional to the overall target performance of the solution.

[0013] Furthermore, the dynamic reallocation strategy includes: The drone was determined to be faulty based on a timeout in the heartbeat signal. The set of targets whose overlap with the original target of the failed UAV in the time window is greater than a preset threshold, or whose spatial distance is less than a preset radius, is identified as a spatiotemporally neighboring target subset; A local model is constructed based on the spatiotemporally nearby target subset and the remaining available UAVs, and the number of solution iterations is limited to ensure that a new allocation scheme is output within the operationally permissible response time.

[0014] Furthermore, when a new drone joins the drone swarm, the method also includes: Identify drones whose current task load exceeds a preset threshold as task migration candidates; combine the newly added drones with the candidate drones to form a local optimization subset; Under the premise of satisfying the constraints, some targets corresponding to the candidate UAVs are migrated to the newly added UAVs; an incremental allocation instruction containing only the migration task is generated, and the remaining UAVs maintain their original flight tasks.

[0015] On the other hand, this paper also provides a UAV swarm weapon target cooperative allocation system under spatiotemporal constraints, the system comprising: The collaborative allocation object definition module is used to define collaborative allocation objects, which include a set of drones, a set of targets, and a set of weapons. The objective function construction module is used to construct the objective function for the collaborative combat scenario. The objective function for the collaborative combat scenario is used to simultaneously optimize the following four objectives: the target survival probability index, the total flight range of the UAV, the mission planning time index for mission completion, and the UAV survival probability index. The constraint definition module is used to define the constraints of the collaborative combat scenario. The constraints include at least time constraints, space constraints, and UAV capability constraints. The coordination and allocation module is used to solve the objective function using a multi-gene population parallel ant colony algorithm. The weapon set, target set, and UAV set are encoded into weapon gene sequences, target gene sequences, and UAV gene sequences, respectively, to generate an initial population. The module performs phased joint optimization by fusing the pheromone guidance mechanism of the ant colony algorithm with the crossover and mutation operation of the genetic algorithm. Under the constraints, the optimal coordination and allocation scheme among the UAVs, weapons, and targets is searched. The dynamic reallocation module is used to trigger a dynamic reallocation strategy if a drone failure is detected during task execution. It only performs local re-optimization on the affected target and the remaining drones to generate a new collaborative allocation scheme.

[0016] Using the above technical solution, the method and system for collaborative allocation of weapons and targets in a UAV swarm under spatiotemporal constraints described in this paper defines a collaborative allocation object, which includes a UAV set, a target set, and a weapon set. Then, a collaborative combat scenario objective function is constructed, which simultaneously optimizes the following four objectives: target survival probability index, total UAV flight range index, mission completion time index, and UAV survival probability index. Further, constraints for the collaborative combat scenario are defined, including at least time constraints, spatial constraints, and UAV capability constraints. Finally, the objective function is solved using a multi-gene population parallel ant colony algorithm, wherein the weapon set, target set, and UAV set are encoded as weapon gene sequences, target gene sequences, and [other gene sequences], respectively. The UAV gene sequence is used to generate an initial population. A phased joint optimization is performed by fusing the pheromone-guided mechanism of ant colony optimization and the crossover and mutation operation of genetic algorithms. Under the given constraints, the optimal collaborative allocation scheme among UAVs, weapons, and targets is searched. During mission execution, if a UAV failure is detected, a dynamic reallocation strategy is triggered, locally re-optimizing only the affected targets and remaining UAVs to generate a new collaborative allocation scheme. This paper establishes a collaborative allocation model for UAVs, weapons, and targets by integrating the traditional independent models of UAV-weapon-target allocation into a unified whole. During the collaborative allocation process, the mutual constraints between the two need to be considered. A mathematical model under spatiotemporal constraints is established, and a collaborative allocation algorithm is designed to solve this problem. The collaborative allocation method can better obtain and optimize UAV mission allocation results, improving the efficiency of UAV collaborative allocation.

[0017] To make the above and other objects, features and advantages of this document more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments or prior art described herein, the accompanying drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this article. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This diagram illustrates the steps of a method for collaborative allocation of weapon targets in a UAV swarm under spatiotemporal constraints, as provided in an embodiment of this paper. Figure 2 This document illustrates a schematic diagram of the parallel gene sequence allocation process in an embodiment of the invention. Figure 3A schematic diagram illustrating the workflow of the multi-gene population parallel ant colony algorithm in the embodiments of this paper is shown. Figure 4 This is a static mission route map of the drone in the simulation example in this article; Figure 5 This is a static mission sequence diagram of the UAV in the simulation example in this article; Figure 6 This is a diagram illustrating the drone task redistribution process in the example presented in this article; Figure 7 This document illustrates a schematic diagram of the framework of a drone swarm weapon target collaborative allocation device under spatiotemporal constraints, as provided in an embodiment of this paper. Figure 8 A schematic diagram of the framework of the computer device provided in the embodiments of this article is shown. Detailed Implementation

[0020] The technical solutions in the embodiments described below will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments described herein, and not all of the embodiments. Based on the embodiments described herein, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this document.

[0021] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings herein are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0022] In existing technologies, the allocation of drone tasks is usually done in stages, without taking into account the constraints and synergy between different coordinating objects.

[0023] To address the aforementioned issues, this paper provides a method for collaborative allocation of weapon targets in UAV swarms under spatiotemporal constraints, which can improve the efficiency of task collaborative allocation among different coordinating objects. Specifically, as shown in the embodiments... Figure 1 As shown, the method may include: S10: Define a collaborative allocation object, which includes a set of drones, a set of targets, and a set of weapons; The embodiments in this specification mainly address the coordination and allocation problem between UAVs, targets, and weapons. In terms of object definition, the first step is to determine the set of defined objects, which includes the parameter information of each object.

[0024] Specifically, drone ensemble: [Settings] For an array of drones, where each drone is represented by a tuple. It indicates. Among them. Indicates the drone serial number. Indicates the location of the drone. Indicates the maximum range of the drone. Indicates the maximum ammunition load of the drone. This indicates the drone's cruising speed.

[0025] Target set: Let The target array, where each target attribute is represented by a tuple. It indicates. Among them. Indicates the target's sequence number. Indicates the target weight. Indicates the location of the target.

[0026] Weapon Collection: Setting This is an array of weapons, where individual weapon attributes are represented by tuples. It indicates. Among them. Indicates the weapon's serial number. This indicates the probability of a weapon damaging a single target.

[0027] S20: Construct a cooperative combat scenario objective function, which is used to simultaneously optimize the following four objectives: target survival probability index, UAV total flight range index, mission completion time index, and UAV survival probability index. In other words, the objective function can be the cost function of a collaborative combat scenario. Generally, the lower the cost, the greater the benefit. By optimizing and updating the parameters in the objective function, the most efficient collaborative allocation scheme can be obtained. The objective function should simultaneously optimize the following four objectives: target survival probability index, total flight range of UAVs, mission completion time index, and UAV survival probability index. When the synergistic relationship of the above four indicators reaches its optimal level, it indicates that the benefit is the highest.

[0028] For example, the objective function of the cooperative combat scenario is expressed by the following formula: ; in, This represents an indicator of the probability of the target surviving. This indicates the total range of the drone. This indicates the timeframe for task planning. This indicates the survival probability index of drones. These represent the weights of the target survival probability, the total flight range of the UAV, the mission planning time, and the UAV survival probability, respectively. .

[0029] Furthermore, in the objective function , , , Calculate using the following formula: ,in, This represents the probability of the target surviving. It is derived from the types and quantities of weapons used to strike a specific target. Furthermore, in calculating the specific result of the target survival probability, it is also necessary to consider the current target's weight value. The final target survival probability is obtained by multiplying the weight by the survival probability. This ensures that weapons with a high probability of destruction prioritize striking heavier targets (such as command posts and radar stations). The survival probability is defined as follows: ,in, Indicates the maximum range of the drone. This indicates that there is no total number of people. Indicates the total number of targets. Indicates the total number of weapons. Indicate weapon type For the target The probability of damage, Indicates assignment to target of The number of type weapons Indicate target j The value or weight of; , The total flight range of a drone represents the total distance traveled by all drones. In calculating this total flight range, the drone's flight range needs to be considered to conserve energy and improve operational efficiency. Therefore, the total flight range of a drone is defined as follows: ;in, Indicates the first A drone from the target fly to target Euclidean distance, Denotes a binary decision variable, if the first... A drone from the target fly to target And to execute the task, the binary decision variable It equals 1, otherwise it equals 0.

[0030] ,in, This refers to the mission planning time, which considers the time required for the slowest drone to complete all tasks, i.e., the total mission duration. Calculating mission planning time can effectively improve mission efficiency, extend the drone swarm's endurance, and support longer-duration missions. The mission planning time is defined as follows: ,in, Indicates the first A drone from the target fly to target Flight time, Indicate target Given the available time window, to ensure mission feasibility, the drone must be deployed no earlier than [date missing]. Reach the destination No later than Complete the task. This indicates the drone's cruising speed.

[0031] ,in, This represents the survival probability of a drone, which means that during mission execution, it is necessary to consider the varying degrees of threat posed to the drone by different targets. By considering the survival probability of drones, the sequence and route of drones performing missions can be planned more effectively. The drone survival probability is defined as follows: ,in, Indicates the maximum range of the drone. Indicates that the drone was targeted The probability of damage.

[0032] By constructing an objective function to achieve multi-objective balance, the search can be made to reduce the probability of survival, shorten the flight range, increase the speed, and improve the survival rate during the collaborative allocation optimization process. By adjusting the weight of the objective function (such as emphasizing "speed" in wartime and "safety" in peacetime), the system can quickly adapt to different combat styles, improving the efficiency and reliability of UAV-target-weapon collaborative allocation.

[0033] S30: Define the constraints for the collaborative combat scenario, which include at least time constraints, space constraints, and UAV capability constraints; This can be understood as follows: after designing the objective function for the collaborative combat scenario, in order to make the allocation algorithm designed in this embodiment closer to the actual battlefield combat environment, it is also necessary to set corresponding constraints on the scenario to ensure that the final UAV mission allocation result is feasible.

[0034] First, time constraints are defined. In this embodiment, time constraints related to adversarial environments need to be considered. In real-world environments, time constraints are a significant factor, including the drone's flight time and the mission's execution time window. Defining these time constraints will ensure the model closely reflects the actual environment. The following are the time constraints established in this paper: This formula represents the time constraint for a drone to travel from its starting point to other targets. Each drone must take off from its starting point, and the time it takes for any drone to execute its mission must be later than the start time of the overall mission plus the time it takes for the drone to travel from its starting point to its current mission target; otherwise, the mission will fail. This formula indicates that the time when the drone arrives at target point j should be within the time window of target point j.

[0035] This formula represents the time constraint for the UAV to reach target point j from target point i. This constraint stipulates that the time from the end of the previous target time window plus the time to fly to the current task must be earlier than the execution time of the current task.

[0036] This formula represents a time sequence constraint, stipulating that the opening time of the task time window cannot be later than the closing time of the task time window.

[0037] in, Indicates the first The drone reached the target. Time, Indicates the specified task start time; Indicates the first A drone from the target fly to target Flight time, Indicates the first A drone flies from the starting point to the target. i Flight time, Indicate target The available time window Represents a binary decision variable. Indicates the first i No more than one drone Complete the task. This indicates that the p-th drone starts from the target. fly to target Binary decision variables on the flight path, This indicates that the p-th drone flies from the starting point to the target. The binary decision variables on the flight path; Besides time constraints, space constraints are also quite important. The algorithm needs to be adjusted according to the corresponding time and space constraints. Therefore, this paper also needs to define the corresponding space constraints, which are as follows: ; ; in, and This indicates that the p-th drone takes off from the starting point and needs to fly back to the starting point after completing its mission.

[0038] This can be understood as the drone starting from the starting point, completing its mission, and then needing to fly back to the starting point. The formula stipulates that the constraint from the drone's takeoff to its final return and landing at the starting point is always 1. Therefore, the drone must start its mission from the starting point and finally return to the starting point.

[0039] Regarding the constraints on drone capabilities, the payload and range of the drone need to be considered, as shown below: ; ; The formulas above all state that the payload and range of each drone cannot exceed their maximum values. Among them, Indicates the first The drone was assigned to the attack target The number of weapons, Indicates the maximum ammunition load of the drone. Indicates the maximum range of the drone. Indicates the first A drone from the target To the target Euclidean distance, This represents a binary decision variable.

[0040] By setting time constraints, space constraints, and UAV capability constraints, the allocation algorithm designed in this embodiment must strictly follow the above time and space constraints, filter out all infeasible solutions, and significantly reduce the search space. These constraints will help make the allocation algorithm designed in this embodiment more closely resemble the real-world UAV adversarial environment and improve the accuracy of collaborative allocation.

[0041] S40: The objective function is solved using a multi-gene population parallel ant colony algorithm, wherein the weapon set, target set, and UAV set are encoded as weapon gene sequences, target gene sequences, and UAV gene sequences, respectively, to generate an initial population. The pheromone guidance mechanism of the ant colony algorithm and the crossover and mutation operation of the genetic algorithm are combined in stages for joint optimization. Under the constraints, the optimal cooperative allocation scheme among UAVs, weapons, and targets is searched.

[0042] In other words, the embodiments in this specification address the collaborative allocation problem involving multiple UAVs, multiple weapons, and multiple targets. They integrate the concept of parallel gene sequences into the ant colony algorithm, combining the optimization capabilities and distributed collaborative advantages of the ant colony algorithm. This fully adapts to the spatiotemporal coupling constraints of the collaborative allocation problem and the limitations of UAV range and payload, effectively improving the effectiveness and real-time performance of the model solution. The overall allocation and algorithm flow is as follows: Figure 2 and Figure 3 As shown.

[0043] Cooperative allocation first requires population initialization, which may include the following steps: The weapon set, target set, and drone set are encoded into weapon gene sequences, target gene sequences, and drone gene sequences, respectively. Random deletion and replacement operations are performed on the target gene sequence, and random sorting is performed on the UAV gene sequence to generate an initial population.

[0044] For example, several target sequences are randomly selected and removed from the sequence. Then, a replacement sequence is randomly selected from the remaining undeleted sequences to fill the vacancy. For instance, if the original target gene sequence is [T1,T2,T3,T4,], and T2 is randomly deleted, then a replacement sequence (e.g., T3) is selected from the remaining sequences. The resulting sequence is then [T1,T3,T3,T4,]. Therefore, by performing local perturbation on the target sequence, we avoid making all individuals identical, thus increasing diversity.

[0045] Furthermore, the specific solution process can be summarized as follows: Based on the aforementioned constraints, the UAV gene sequences are optimized to determine the mission access order for each UAV. The assignment relationship between fixed UAVs and targets is established, and the UAV gene sequence and the target gene sequence are combined into a multidimensional gene sequence. Based on the aforementioned multidimensional gene sequence, and in conjunction with the weapon parameters of each weapon, the weapon gene sequence is optimized; The optimized parallel gene sequences are fused together, and the candidate solutions are sorted using the objective function to perform Pareto non-dominated sorting. Elite solutions are then selected, and the pheromones are updated based on the elite solutions. Each generation of elite solutions is directly passed to the next generation of the population, and the population is updated through crossover and mutation operations. This process is iterated and optimized until the convergence condition is met.

[0046] When optimizing the UAV gene sequence, it is necessary to comprehensively consider the UAV's range constraints, mission execution time window, and target priority (i.e. weight). Heuristic rules based on path planning can be used to adjust the access order of targets in the sequence. For example, by constructing a task priority matrix and a UAV capability matching matrix, high-priority targets with high matching degree with UAV capabilities can be arranged in reasonable access positions to reduce invalid flight paths of UAVs.

[0047] During the combination of multidimensional gene sequences, it is necessary to ensure that the length of the UAV gene sequence is consistent with that of the target gene sequence, and that each UAV and target form a one-to-one assignment relationship to avoid conflicting assignments such as one-to-many or many-to-one. When optimizing the weapon gene sequence, a weapon-target matching degree evaluation mechanism is established based on the weapon parameters, such as the probability of damage to the target and the target type. The optimal weapon type and quantity are matched for each target-UAV combination using a greedy algorithm or local search strategy. For example, for targets such as radar, armored vehicles, and command posts, weapons with different damage capabilities are selected, while ensuring that the number of weapons does not exceed the payload limit of the UAV.

[0048] When performing Pareto non-dominated ranking, multiple indicators such as target survival probability, total UAV flight range, mission completion planning time, and UAV survival probability are used as evaluation metrics. Candidate solutions are divided into different non-dominated levels, and elite solutions at the forefront of the optimal solution are selected. These elite solutions not only perform well in each target indicator but also provide high-quality feedback for pheromone updates. Pheromones are updated in the following way: The pheromone is updated in the following manner: ; ; in, This represents the pheromone along the execution path from target i to target j; Indicates the evaporation rate. Indicates the initial pheromone evaporation rate. This represents the final pheromone evaporation rate, and satisfies 0 < < <1; t Represents iterative algebra, >0 represents the decay rate parameter, used to control the transition speed of the evaporation rate from the exploration stage to the development stage.

[0049] Furthermore, the multi-gene population parallel ant colony algorithm performs Pareto non-dominated sorting on candidate solutions in each iteration, retains the non-dominated solutions as the elite solution set, and performs a pheromone enhancement operation based on the elite solution set, wherein the pheromone increment Δ It is inversely proportional to the overall target performance of the solution.

[0050] The pheromone update adopts an elite solution-guided mechanism, which increases the pheromone concentration for paths or assignment schemes corresponding to elite solutions, while performing appropriate pheromone evaporation processing on other non-elite solutions to balance the algorithm's exploration and development capabilities.

[0051] Crossover operations can be performed using partial mapping crossover or sequential crossover, exchanging fragments of the multidimensional gene sequences of selected parent individuals to generate offspring individuals with new gene combinations; mutation operations are achieved by randomly exchanging the positions of elements in the gene sequence or replacing specific gene fragments, further enhancing the diversity of the population and preventing the algorithm from getting trapped in local optima.

[0052] During the iterative optimization process, the convergence status is determined by monitoring the changes in the objective function value of elite solutions over multiple generations. When the improvement of elite solutions within a preset number of generations is less than a set threshold, or when the number of iterations reaches a preset threshold, the iteration is terminated and the current optimal collaborative allocation scheme is output.

[0053] In other words, this manual employs a phased optimization strategy. First, it performs individual optimization, locally optimizing the execution order of the UAV gene sequences to ensure that the UAV cluster has superior basic performance when running independently. Then, it performs joint optimization based on the set of individual optimizations, treating all parallel gene sequences as an organic whole. By constructing a multi-objective optimization function, it comprehensively considers multiple constraints between tasks and uses a multi-gene population parallel ant colony algorithm for global optimization. This achieves gradual optimization from local to global, effectively avoiding the inefficiency caused by an excessively large search space when directly performing global optimization.

[0054] S50: If a drone failure is detected during mission execution, a dynamic reallocation strategy is triggered, which only performs local re-optimization on the affected target and the remaining drones to generate a new collaborative allocation scheme.

[0055] In other words, the embodiments of this specification, through real-time monitoring of the drone status, can promptly identify failed drones and update and adjust the targets assigned to those failed drones, thereby ensuring that all missions against the targets are completed and improving the reliability of mission completion. Specifically, the following process can be retained: The drone was determined to be faulty based on a timeout in the heartbeat signal. The set of targets whose overlap with the original target of the failed UAV in the time window is greater than a preset threshold, or whose spatial distance is less than a preset radius, is identified as a spatiotemporally neighboring target subset; A local model is constructed based on the spatiotemporally nearby target subset and the remaining available UAVs, and the number of solution iterations is limited to ensure that a new allocation scheme is output within the operationally permissible response time.

[0056] In other words, the system determines whether a drone has failed based on feedback signals between the drone and the control center. If there is no feedback between the drone and the control center for an extended period, it indicates that the drone has failed. In this case, partial updates can be performed on the targets assigned to the drone to ensure that other normally functioning drones can take over the tasks previously handled by the failed drone. First, the identification information, current task execution progress, and target set of the failed drones are obtained. Simultaneously, real-time status parameters such as the number of remaining available drones, their current location, remaining battery power, and load capacity are calculated. Next, the target tasks corresponding to the failed drones are extracted from the original collaborative allocation scheme and marked as a task set to be reassigned. When constructing the local model, the original allocation scheme for unaffected drones is retained first, and target tasks in the spatiotemporally adjacent target subset are reassigned only to avoid excessive disturbance to the overall collaborative framework due to local adjustments. Then, a simplified multi-gene population parallel ant colony algorithm is used to quickly solve this local model. Gene sequence recombination and optimization are performed only on the task set to be reassigned and the remaining available drones. This significantly shortens the response time of dynamic adjustments while ensuring the effectiveness of the allocation scheme, ensuring the continuity and stability of task execution, and ensuring that the new allocation scheme is both efficient and economical.

[0057] In one embodiment of this specification, when a new drone is added, the collective burden of the current drone swarm can be reduced, and other drones can be quickly and efficiently assisted, thereby improving the overall mission efficiency. Specifically, this can be achieved by: Identify drones whose current task load exceeds a preset threshold as task migration candidates; combine the newly added drones with the candidate drones to form a local optimization subset; Under the premise of satisfying the constraints, some targets corresponding to the candidate UAVs are migrated to the newly added UAVs; an incremental allocation instruction containing only the migration task is generated, and the remaining UAVs maintain their original flight tasks.

[0058] The task load can be determined from multiple dimensions, including the drone's remaining battery power, the duration of executed tasks, and the complexity of the target tasks (such as time windows and the number of targets). Drones with higher load pressure but relatively lower task urgency are prioritized to avoid interfering with the execution of high-priority tasks. For the local optimization subset consisting of new drones and candidate drones, an evaluation model for task migration needs to be established. This model can aim to minimize the overall energy consumption and the total task completion time after migration, while also considering the load capacity constraints of new drones and the spatiotemporal constraints of target tasks. In the specific task migration process, the target tasks of candidate drones are first prioritized, and lower-priority tasks with a high spatiotemporal match with the current location of the new drone are migrated first. After migration, the task execution paths before and after migration are locally planned and adjusted to ensure that the new drone can execute the migration task on the optimal path, while the remaining task paths of candidate drones are also optimized accordingly. Ultimately, this achieves a balanced distribution of task load within the local subset and an improvement in overall execution efficiency.

[0059] The spatiotemporal constraints-constrained UAV swarm weapon-target collaborative allocation method provided in this specification addresses the weapon-target collaborative allocation problem. It proposes a multi-population genetic ant colony optimization (MPGACO) algorithm by integrating the population update method of a genetic algorithm into an ant colony algorithm. The parallel gene sequence strategy of the genetic algorithm enables better collaborative allocation of multiple targets. Integrating this strategy into the ant colony algorithm leverages the algorithm's fast convergence speed to improve efficiency. Simultaneously, the strong global search capability of the elite genetic strategy helps the algorithm escape local optima and find the global optimum. This method considers the constraints and synergy between the two during modeling and allocation, achieving better UAV mission allocation results in resource-constrained environments with high target engagement requirements, while also meeting real-time requirements during combat.

[0060] In a specific example, the method proposed in this specification is simulated using the MATLAB experimental platform. Regarding the specific implementation, to facilitate algorithm verification and analysis, the following assumptions are made: 1. The positions of all targets are known, and the position change occurs at most once; the changed position is also assumed to be known. 2. All targets are reachable, and all flight paths are flyable. 3. In this experiment, the UAV flight trajectory is not considered; all flight paths are assumed to be straight lines. Table 1 defines the experimental settings in the implementation process. Each target has its own time window, which remains unchanged during task execution. The flight time between each task is determined by the straight-line distance between the two task points and the UAV's flight speed, which is assumed to be a constant value. The weight value of each target point is randomly generated during the initialization phase, representing the target's importance.

[0061] Table 1 Object Parameter Settings

[0062] The method proposed in the embodiments of this specification combines a parallel genetic strategy with an ant colony algorithm to design a collaborative allocation algorithm for multiple objects. The algorithm's performance was verified using a MATLAB simulation platform; detailed algorithm parameter settings are shown in Table 2.

[0063] Table 2 Algorithm Parameter Settings

[0064] Based on the model and algorithm settings mentioned above, the algorithm proposed in this invention first generates an initial allocation result based on the currently known information. Then, by having 15 drones execute tasks on targets within the area, a time-series task allocation result is obtained according to the previously designed objective function and constraints. This allocation result serves as the initial task execution order for the drone swarm. During the drone's task execution, drone damage may occur. If the environment changes, the simulation algorithm will respond promptly to the environmental changes and provide the allocation result after the environmental changes.

[0065] Figure 4 This is the initial mission route map for 15 drones. It's possible that some targets may not be assigned, or multiple drones may engage the same target. This is because each target has a weight during initialization, and weapons have different damage probabilities against targets. The algorithm prioritizes important targets and discards those with lower weights. Furthermore, since drones have limited payload capacity, when the ammunition required for a single target exceeds the drone's carrying capacity, the algorithm will comprehensively allocate resources based on target weight and weapon damage probability, ultimately generating the drones' mission execution routes.

[0066] Figure 5This is a time sequence diagram for 15 drones. Each row in the diagram represents the time sequence of a drone's mission execution. Blue represents the drone's flight time, yellow represents the waiting time, and red represents the mission execution time. Since each mission has a time window, the constraints stipulate that the drone must reach the target point and execute the mission within the time window. Therefore, the time sequence diagram shows that drones may have a certain waiting time after reaching the target point, and the number of missions executed by different drones varies.

[0067] As time progresses, the simulation algorithm will reallocate the task allocation results based on changes in the environment. Figure 6 The image shows the entire process of drone redistribution. Among them, Figure 6 (a) represents the initial task assignment result for the drone. As time progresses, the drone may suffer casualties during the execution of its mission. Figure 6 In diagram (b), the red paths represent damaged drones. UAV3, UAV7, and UAV12 are damaged and unable to continue their missions, and their assigned targets are also missing weapon engagements. Because the redistribution process does not redistribute all missions, only a portion of the drones are affected, thus reducing the computational load during the redistribution process. Figure 6 In section (c), four of the drones were reassigned, and the results of the reassignment are shown. The black dashed lines represent the drones that were not reassigned. Figure 6 The middle (d) shows the overall allocation result after redistribution.

[0068] To comprehensively evaluate the performance of the proposed method, this specification introduces genetic algorithms, ant colony algorithms, and parallel elite ant colony algorithms as comparative methods in its embodiments. Experimental results show that the algorithm proposed in this specification's embodiments performs well in terms of final cost and computational efficiency, improving the response speed of UAV swarms in dynamic combat environments.

[0069] Based on the aforementioned method for collaborative allocation of weapon targets in UAV swarms under spatiotemporal constraints, this paper also provides a system for collaborative allocation of weapon targets in UAV swarms under spatiotemporal constraints, such as... Figure 7 As shown, the system includes: The collaborative allocation object definition module 710 is used to define collaborative allocation objects, which include a set of drones, a set of targets, and a set of weapons. The objective function construction module 720 is used to construct the objective function for the collaborative combat scenario. The objective function for the collaborative combat scenario is used to simultaneously optimize the following four objectives: the target survival probability index, the total flight range index of the UAV, the mission completion planning time index, and the UAV survival probability index. The constraint definition module 730 is used to define the constraints of the collaborative combat scenario, and the constraints include at least time constraints, space constraints and UAV capability constraints. The coordination and allocation module 740 is used to solve the objective function using a multi-gene population parallel ant colony algorithm. The weapon set, target set, and UAV set are encoded into weapon gene sequences, target gene sequences, and UAV gene sequences, respectively, to generate an initial population. The module performs phased joint optimization by fusing the pheromone guidance mechanism of the ant colony algorithm with the crossover and mutation operation of the genetic algorithm. Under the constraints, the optimal coordinated allocation scheme among the UAVs, weapons, and targets is searched. The dynamic reallocation module 750 is used to trigger a dynamic reallocation strategy if a drone failure is detected during task execution. It only performs local re-optimization on the affected target and the remaining drones to generate a new collaborative allocation scheme.

[0070] This embodiment provides a computer device, the internal structure of which can be shown in the following diagram. Figure 8 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection.

[0071] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0072] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0073] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0074] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0075] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. The databases involved in the embodiments provided in this application can include at least one of relational and non-relational databases. Non-relational databases can include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application can be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0076] In the embodiments provided herein, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the couplings or direct couplings or communication connections shown or discussed may be indirect couplings or communication connections through some interfaces, devices, or units, or they may be electrical, mechanical, or other forms of connection.

[0077] This document uses specific embodiments to illustrate the principles and implementation methods of this document. The descriptions of the embodiments above are only for the purpose of helping to understand the methods and core ideas of this document. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this document. Therefore, the content of this specification should not be construed as a limitation of this document.

Claims

1. A method for collaborative allocation of weapon targets in a UAV swarm under spatiotemporal constraints, characterized in that, The method includes: S10: Define a collaborative allocation object, which includes a set of drones, a set of targets, and a set of weapons; S20: Construct a cooperative combat scenario objective function, which is used to simultaneously optimize the following four objectives: target survival probability index, UAV total flight range index, mission completion time index, and UAV survival probability index. S30: Define the constraints for the collaborative combat scenario, which include at least time constraints, space constraints, and UAV capability constraints; S40: The objective function is solved using a multi-gene population parallel ant colony algorithm, wherein the weapon set, target set, and UAV set are encoded as weapon gene sequences, target gene sequences, and UAV gene sequences, respectively, to generate an initial population. The pheromone guidance mechanism of the ant colony algorithm and the crossover and mutation operation of the genetic algorithm are combined in stages for joint optimization. Under the constraints, the optimal cooperative allocation scheme among UAVs, weapons, and targets is searched. S50: If a drone failure is detected during mission execution, a dynamic reallocation strategy is triggered, which only performs local re-optimization on the affected target and the remaining drones to generate a new collaborative allocation scheme.

2. The method according to claim 1, characterized in that, The objective function for the cooperative combat scenario is expressed by the following formula: ； in, This represents an indicator of the probability of the target surviving. This indicates the total range of the drone. This indicates the time frame for task planning. This indicates the survival probability index of drones. These represent the weights of the target survival probability, the total flight range of the UAV, the mission planning time, and the UAV survival probability, respectively. ; , Indicates the probability of the target surviving. , Indicates the maximum range of the drone. This indicates that there is no total number of people. Indicates the total number of targets. Indicates the total number of weapons. Indicate weapon type For the target The probability of damage, Indicates assignment to target of The number of type weapons Indicate target j The weights; , Indicates the total flight range of the drone. ; Indicates the first A drone from the target To the target Euclidean distance, Represents a binary decision variable; , Indicates the task planning time. , Indicates the first A drone from the target fly to target Flight time, Indicate target The available time window Indicates the drone's cruising speed; , Indicates the survival probability of a drone. , Indicates the maximum range of the drone. Indicates that the drone was targeted The probability of damage.

3. The method according to claim 1, characterized in that, The constraints of the collaborative combat scenario are expressed by the following formula: Time constraints: ； ； ； ； in, Indicates the first The drone reached the target. Time, Indicates the specified task start time; Indicates the first A drone from the target fly to target Flight time, Indicates the first A drone flies from the starting point to the target. i Flight time, Indicate target The available time window Represents a binary decision variable. Indicates the first i No more than one drone Complete the task. This indicates that the p-th drone starts from the target. fly to target Binary decision variables on the flight path, This indicates that the p-th drone flies from the starting point to the target. The binary decision variables on the flight path; Spatial constraints: ； ； in, and This means that the p-th drone takes off from the starting point and needs to fly back to the starting point after completing its mission; Unmanned aerial vehicle (UAV) capability constraints: ； ； in, Indicates the first The drone was assigned to the attack target The number of weapons, Indicates the maximum ammunition load of the drone. Indicates the maximum range of the drone. Indicates the first A drone from the target To the target Euclidean distance, This represents a binary decision variable.

4. The method according to claim 1, characterized in that, The initial population in step S4 is obtained through the following steps: The weapon set, target set, and drone set are encoded into weapon gene sequences, target gene sequences, and drone gene sequences, respectively. Random deletion and replacement operations are performed on the target gene sequence, and random sorting is performed on the UAV gene sequence to generate an initial population.

5. The method according to claim 1, characterized in that, Step S4 includes: Based on the aforementioned constraints, the UAV gene sequences are optimized to determine the mission access order for each UAV. The assignment relationship between fixed UAVs and targets is established, and the UAV gene sequence and the target gene sequence are combined into a multidimensional gene sequence. Based on the aforementioned multidimensional gene sequence, and in conjunction with the weapon parameters of each weapon, the weapon gene sequence is optimized; The optimized parallel gene sequences are fused together, and the candidate solutions are sorted using the objective function to perform Pareto non-dominated sorting. Elite solutions are then selected, and the pheromones are updated based on the elite solutions. Each generation of elite solutions is directly passed to the next generation of the population, and the population is updated through crossover and mutation operations. This process is iterated and optimized until the convergence condition is met.

6. The method according to claim 1, characterized in that, The pheromone is updated in the following manner: ； ； in, This represents the pheromone along the execution path from target i to target j; Indicates the evaporation rate. Indicates the initial pheromone evaporation rate. This represents the final pheromone evaporation rate, and satisfies 0 < < <1; t Represents iterative algebra, >0 represents the decay rate parameter, used to control the transition speed of the evaporation rate from the exploration stage to the development stage.

7. The method according to claim 6, characterized in that, The multi-gene population parallel ant colony algorithm performs Pareto non-dominated sorting on candidate solutions in each iteration, retains the non-dominated solutions as the elite solution set, and performs pheromone enhancement operation based on the elite solution set, wherein the pheromone increment Δ It is inversely proportional to the overall target performance of the solution.

8. The method according to claim 1, characterized in that, The dynamic reallocation strategy includes: The drone was determined to be faulty based on a timeout in the heartbeat signal. The set of targets whose overlap with the original target of the failed UAV in the time window is greater than a preset threshold, or whose spatial distance is less than a preset radius, is identified as a spatiotemporally neighboring target subset; A local model is constructed based on the spatiotemporally nearby target subset and the remaining available UAVs, and the number of solution iterations is limited to ensure that a new allocation scheme is output within the operationally permissible response time.

9. The method according to claim 1, characterized in that, When a new drone joins the drone swarm, the method further includes: Identify drones whose current task load exceeds a preset threshold as task migration candidates; combine the newly added drones with the candidate drones to form a local optimization subset; Under the premise of satisfying the constraints, some targets corresponding to the candidate UAVs are migrated to the newly added UAVs; an incremental allocation instruction containing only the migration task is generated, and the remaining UAVs maintain their original flight tasks.

10. A collaborative weapon target allocation system for unmanned aerial vehicle (UAV) swarms under spatiotemporal constraints, characterized in that, The system includes: The collaborative allocation object definition module is used to define collaborative allocation objects, which include a set of drones, a set of targets, and a set of weapons. The objective function construction module is used to construct the objective function for the collaborative combat scenario. The objective function for the collaborative combat scenario is used to simultaneously optimize the following four objectives: the target survival probability index, the total flight range of the UAV, the mission planning time index for mission completion, and the UAV survival probability index. The constraint definition module is used to define the constraints of the collaborative combat scenario. The constraints include at least time constraints, space constraints, and UAV capability constraints. The coordination and allocation module is used to solve the objective function using a multi-gene population parallel ant colony algorithm. The weapon set, target set, and UAV set are encoded into weapon gene sequences, target gene sequences, and UAV gene sequences, respectively, to generate an initial population. The module performs phased joint optimization by fusing the pheromone guidance mechanism of the ant colony algorithm with the crossover and mutation operation of the genetic algorithm. Under the constraints, the optimal coordination and allocation scheme among the UAVs, weapons, and targets is searched. The dynamic reallocation module is used to trigger a dynamic reallocation strategy if a drone failure is detected during task execution. It only performs local re-optimization on the affected target and the remaining drones to generate a new collaborative allocation scheme.

Images