Aircraft array position reconstruction method based on local conflict resolution auction method

Abstract

The invention discloses an aircraft array position reconstruction method based on a local conflict resolution auction method. The method comprises the following steps: 1) modeling a multi-unmanned aerial vehicle system, and defining an array position distribution matrix, a target function and a distribution constraint; 2) constructing an index performance function; 3) designing a local conflict resolution auction-based array location allocation algorithm, and enabling each unmanned aerial vehicle to independently calculate an optimal target based on a performance evaluation result of the unmanned aerial vehicle; and 4) recording a global array position distribution result and evaluating validity. According to the method, the problems of poor robustness of a centralized scheme, insufficient adaptability to aircraft faults and the like in the existing unmanned aerial vehicle formation position distribution are solved.

Description

Technical Field

[0001] This invention relates to the field of aircraft formation control and position allocation, specifically to an aircraft position reconstruction method based on the local conflict resolution auction method. Background Technology

[0002] Multi-UAV swarms are widely used in reconnaissance, patrol, and communication relay missions, often facing interference from complex environmental conditions and unforeseen events, such as electromagnetic degradation or individual node failures. These extreme situations can cause some aircraft to malfunction, thereby disrupting the swarm structure, reducing overall collaborative efficiency, and even leading to mission failure. Therefore, UAV swarm systems with rapid response and adaptive reconfiguration capabilities in dynamic environments have become a crucial area for urgent development.

[0003] Most array position allocation methods rely on optimization algorithms based on global information, but they suffer from high communication overhead and weak robustness, making it difficult to meet the flexibility and reliability requirements of real-world scenarios. Therefore, there is an urgent need to propose an array position allocation mechanism with distributed decision-making, real-time adaptability, and fault tolerance capabilities to complete space deployment and reorganization under spacecraft failure conditions. Summary of the Invention

[0004] The purpose of this invention is to provide an aircraft position reconfiguration method based on the local conflict resolution auction method, which solves the problems of poor robustness and insufficient adaptability to aircraft failures in the existing centralized scheme for UAV formation position allocation.

[0005] To achieve the above objectives, this invention provides an aircraft position reconfiguration method based on the local conflict resolution auction method, comprising: 1) multi-UAV system modeling, defining position allocation matrix, objective function and allocation constraints; 2) constructing an index performance function; 3) designing a position allocation algorithm based on local conflict resolution auction, enabling each UAV to independently calculate its optimal objective based on its own performance evaluation results; 4) recording and evaluating the effectiveness of global position allocation results.

[0006] The above-mentioned aircraft position reconfiguration method based on the local conflict resolution auction method includes the following allocation constraints in step 1): ① Each valid UAV must be assigned to a target position; ② Each target position can only be assigned to one UAV.

[0007]

[0008] The effective set of dynamic changes of UAVs is denoted as V = {v1, v2, ..., v...} N}, v i The i-th drone that is still capable of performing the mission at the current moment is represented by N, where N is the number of active drones at the current moment. This indicates that the effective drone vi Assigned to target position t j The efficiency value; a ij ∈[0,1] represents the corresponding binary decision variables; the target matrix set is defined as T={t1,...,t M}, N≤M; the matrix for assigning positions is defined as A=[a ij ] N×M The objective function reflects the expected effect of the overall array allocation based on the contributions of each individual UAV.

[0009] The above-mentioned aircraft position reconstruction method based on the local conflict resolution auction method, wherein step 2) includes: defining the normalized distance performance:

[0010]

[0011] In the formula, r ij Indicates drone v i With target position t j The distance between them; This represents the average distance between all drones and the target array.

[0012] Define heading alignment effectiveness:

[0013]

[0014] In the formula, the average angular deviation is denoted as Angle term defined as Indicates drone v i The velocity vector;

[0015] The importance and effectiveness of the formation positions are as follows:

[0016]

[0017] in, The attenuation coefficient is adjustable. Indicates the index of the target array position;

[0018] Overall performance function:

[0019]

[0020] In the formula, the weighting coefficients satisfy the constraints. and i = 1, 2, ..., N, where N is the number of active UAVs at the current moment; j = 1, 2, ..., M, where M is the number of target positions.

[0021] The above-mentioned aircraft position reconstruction method based on the local conflict resolution auction method includes, in step 3), the following steps: 3.1) Calculating target effectiveness index: Each effective UAV calculates its effectiveness value for all target positions based on its current state information. The effectiveness function integrates relative distance, heading angle, and the importance of the target position to obtain the comprehensive effectiveness vector of the UAV; 3.2) Selecting the optimal target and updating the bid based on local benefits: Each effective UAV selects the target corresponding to the maximum net profit as the current bidding target based on the difference between effectiveness and current price, and records the corresponding profit value to achieve local optimal allocation; 3.3) Broadcasting its own bid and receiving neighbor bid information: Broadcasting its current bidding target and corresponding profit to neighbor nodes through the communication link, and receiving the optimal target and corresponding profit from neighbor nodes; 3.4) Forming a local bidding set: Integrating its own and neighbor bidding information to form a local "node-target-profit" set, identifying the neighbor set with the same target selection as itself, and constituting the conflict set on the current target; 3.5 3.1) Determine if there is a target conflict: If the conflict set is empty, the current target uniquely corresponds to the current drone, which is considered to be without conflict. Directly confirm the current allocation result and proceed to step 3.8); If there is a conflict, proceed to step 3.6); 3.6) Compare its own and its neighbors' revenue values ​​to determine if it has the highest bid: The drone compares its current revenue with the revenue of other neighbors in the conflict set. Based on the revenue comparison result, it determines whether it is the highest-revenue party. If its own revenue is the highest in the conflict set, it retains the current bidding plan and proceeds to step 3.8); otherwise, proceed to step 3.7) to adjust the bid; 3.7) Update the target price and recalculate the target selection: If the bid of this node is not advantageous, the target price is adjusted, and the current target selection and corresponding revenue are updated based on the new price. Return to step 3.2) to restart a new round of bidding; 3.8) Output the final position allocation result: When a valid drone completes the position selection and there is no conflict, the drone confirms the current allocation result and completes the position allocation for that drone.

[0022] The above-mentioned aircraft position reconfiguration method based on the local conflict resolution auction method, wherein step 4) includes: introducing a binary decision variable a to describe the global position allocation scheme at any given time. ij The final global matrix A is:

[0023]

[0024] Assuming a virtual central node exists, it uniformly collects and performs posterior evaluation of the position allocation results for each UAV. This evaluation not only outputs the global position allocation matrix but also, based on the constructed comprehensive performance function,... Calculate the overall average benefit index J assign As a quantitative evaluation standard for task matching quality, J assignThe specific definition is:

[0025]

[0026] In the formula, For drones v i Assigned to target position t j The overall performance value, J assign It reflects the average level of efficiency across all current distribution relationships.

[0027] Compared with the prior art, the beneficial technical effects of the present invention are:

[0028] (1) This invention transforms the position reconstruction problem into an aircraft-position target allocation problem by defining the position allocation matrix and the comprehensive performance function, and constructs an auction game algorithm under dynamic topology and an adaptive evaluation index to provide a problem analysis framework for algorithm analysis and evaluation.

[0029] (2) This invention designs a weighted efficiency function to guide the dynamic adjustment of the array position in terms of distance, heading and importance, proposes a local price compensation mechanism, realizes the decentralized resolution of conflicts through neighbor information, and can still quickly complete the array position reconstruction under the condition that node communication is disturbed or fails.

[0030] (3) Through local communication and autonomous decision-making mechanisms, this invention enables each aircraft to effectively avoid conflicts and adapt to dynamic changes such as node failures during the allocation of positions, thereby achieving coordinated position adjustment under local perception and limited communication conditions, and enhancing the mission sustainability and survivability of the overall formation system. Attached Figure Description

[0031] The aircraft position reconstruction method based on the local conflict resolution auction method of the present invention is given by the following embodiments and figures.

[0032] Figure 1 This is a flowchart of the aircraft position reconstruction method based on the local conflict resolution auction method according to an embodiment of the present invention.

[0033] Figure 2 This is a schematic diagram of a preset formation configuration in an embodiment of the present invention.

[0034] Figure 3 This is a schematic diagram illustrating the construction of the indicator performance function in an embodiment of the present invention.

[0035] Figure 4 This is a flowchart of the array position allocation algorithm based on local conflict resolution auction in an embodiment of the present invention.

[0036] Figure 5 This is a schematic diagram of the first round of auction results for formation position allocation in an embodiment of the present invention.

[0037] Figure 6 This is a schematic diagram of the second round of auction results for formation position allocation in an embodiment of the present invention.

[0038] Figure 7 This is a schematic diagram of the third round of auction results for formation position allocation in an embodiment of the present invention.

[0039] Figure 8 This is a schematic diagram of the fourth round of auction results for formation position allocation in an embodiment of the present invention. Detailed Implementation

[0040] The following will combine Figures 1 to 8 The method for reconfiguring aircraft positions based on the local conflict resolution auction method of the present invention will be described in further detail.

[0041] To address the shortcomings of existing multi-UAV formation strategy fault tolerance in dynamic mission environments and to ensure rapid formation reconfiguration under unforeseen events, this invention proposes an aircraft formation reconfiguration method based on the local conflict resolution auction method.

[0042] The spacecraft position reconstruction method based on the local conflict resolution auction method of the present invention includes:

[0043] 1) Multi-UAV system modeling: Define the position allocation matrix, objective function and allocation constraints for the matching relationship between aircraft (UAVs) and target positions in the formation;

[0044] The core objective of formation reconfiguration is to optimize cluster task performance while adhering to strict allocation constraints.

[0045] Specific allocation constraints include: ① Each valid UAV must be assigned to a target position; ② Each target position can only be assigned to one UAV.

[0046]

[0047] The effective set of dynamic changes of UAVs is denoted as V = {v1, v2, ..., v...} N}, v i The i-th drone that is still capable of performing the mission at the current moment is represented by N, which is the number of active drones at the current moment. Due to the possibility of interference from factors such as sudden events, communication interruptions or physical damage, members in working condition are always selected from the entire drone set. This indicates that the effective drone v i Assigned to target position t j The efficiency value is a dimensionless indicator reflecting the quality of allocation decisions; the higher the value, the better the allocation plan is under the current circumstances. ij∈[0,1] represents the corresponding binary decision variables; the target matrix set is defined as T={t1,...,t M It should be noted that due to the impact of unforeseen events such as malfunctions, the number of effective UAVs N varies over time, but never exceeds the number of target positions, i.e., N≤M. This dynamic adaptability is a key requirement for achieving robustness and flexible formation reconfiguration in real combat environments; the position allocation matrix is ​​defined as A=[a ij ] N×M The objective function reflects the expected effect of the overall array allocation based on the contributions of each individual UAV.

[0048] 2) Construction of indicator performance function: In order to measure the quality of different UAVs assigned to different positions, a comprehensive performance function integrating three types of indicators is constructed, including relative distance, heading consistency and position importance;

[0049] To address the target allocation problem and transform actual combat requirements into quantifiable allocation indicators, this invention proposes a multi-indicator fusion performance function that comprehensively considers three key operational factors: relative distance, heading alignment accuracy, and position importance.

[0050] Range performance metrics: To reflect the effective proximity between the UAV and the target array, the normalized range performance is defined as follows:

[0051]

[0052] In the formula, r ij Indicates drone v i With target position t j The distance between them; The ratio ||r| represents the average distance between all effective UAVs and the target array. ij ‖ / δ r The distances under different mission scenarios were normalized to ensure the scale invariance of the performance function; the exponential form compresses the distance performance value within the interval (0,1), so that the smaller the distance, the less maneuvering is required for the UAV to reach the target position, and the higher the performance.

[0053] Heading alignment effectiveness index: To quantify the degree of alignment between the UAV's current direction of motion and the line-of-sight direction of the target position, the heading alignment effectiveness is defined as follows:

[0054]

[0055] In the formula, the average angular deviation is denoted as Angle term defined as The smaller the angle, the closer the current course is to the target direction, the smaller the adjustment required, and the higher the course efficiency. Indicates drone v i The velocity vector; the exponential form is also used to compress heading alignment performance values ​​to a comparable range and to emphasize heading smoothness;

[0056] Position Importance Performance Index: To reflect the differences in the importance of target positions, the following exponential decay function is defined:

[0057]

[0058] in, The attenuation coefficient is adjustable. This represents the index of the target position; this function assumes that the smaller the target position number, the higher its importance in the formation, and therefore the higher its priority.

[0059] Comprehensive Performance Function: The above three indicators are weighted and integrated to form a unified comprehensive performance function, as follows:

[0060]

[0061] In the formula, the weighting coefficients satisfy the constraints. and This multi-index performance function serves as an evaluation criterion in position allocation optimization, driving UAVs to implement reasonable target position reallocation strategies under dynamic formation configurations and changing mission requirements. This is achieved by integrating distance, heading, and position importance into a scalar index. This invention ensures that the auction mechanism directly optimizes task-related trade-offs;

[0062] 3) Design a position allocation algorithm based on local conflict resolution auction, so that each UAV can independently calculate its optimal target based on its own performance evaluation results;

[0063] Inspired by the consensus auction algorithm, this invention proposes a position allocation algorithm based on a distributed auction mechanism (i.e., a position allocation algorithm based on local conflict resolution auction). In this algorithm, each active drone locally calculates its performance for each target position, obtains the drone's overall performance vector, and infers a globally conflict-free position allocation result through an iterative bidding process. Furthermore, the drone does not need to obtain complete performance information from its neighbors; it can converge to a consistent position allocation simply by minimizing the exchange of bidding information.

[0064] Initial target selection: Each UAV selects its current optimal target based on its local performance vector (i.e., based on its overall performance vector).

[0065]

[0066] In the formula, and γ j These represent the current assignment of the drone to the target position t. j The overall efficiency, local price, and optimal net profit, therefore, j * Corresponding benefits This represents the current maximum net profit;

[0067] Conflict resolution mechanism: The core of the distributed auction algorithm lies in resolving potential conflicts through a revenue-driven price adjustment mechanism; each drone does not need to publicly disclose its complete performance matrix, but only needs to broadcast its current bid pair (j) to its neighbors. * ,γ * When multiple neighboring drones select the same target position, a conflict is considered to exist, and the conflict set is defined as follows: in For the target position selection of neighbor m; the drone will achieve its optimal benefit γ * Maximum benefit of other nodes in conflict concentration Compare, if γ * <γ max Then the drone updates the price of the current target.

[0068]

[0069] In the formula, k is the discrete iteration number of the current auction algorithm. For target j * The quote after the k-th update and the quote after the (k+1)-th update. ò is a small positive number used to ensure the convergence of the algorithm; if there is no conflict, or the drone has the highest reward in the conflict set, then it confirms the current allocation result.

[0070] Unlike many auction algorithms that rely on centralized management or a global performance matrix, the method proposed in this invention completes conflict detection and price adjustment entirely within the local neighborhood, which greatly reduces communication overhead and avoids dependence on global consistency.

[0071] In summary, the array position allocation algorithm based on the distributed auction mechanism proposed in this invention enables agents to achieve conflict-free target array position allocation under limited communication conditions through a local price adjustment mechanism, providing an effective means for achieving robust, scalable, and fully distributed array reconfiguration under critical task constraints.

[0072] Specifically, it includes the following steps:

[0073] 3.1) Calculate the target effectiveness index: Based on the current state information, each effective UAV calculates its effectiveness value for all target positions. The effectiveness function integrates relative distance, heading angle and the importance of target positions to obtain the local comprehensive effectiveness vector (i.e., the comprehensive effectiveness vector of the UAV).

[0074] 3.2) Select the optimal target and update the bid based on local benefits: Each effective UAV selects the target with the maximum net profit as the current bidding target based on the difference between its effectiveness and the current price, and records the corresponding profit value to achieve local optimal allocation;

[0075] 3.3) Broadcast its own bid and receive neighbor bid information: Broadcast its current bidding target and corresponding profit to neighbor nodes through the communication link, and receive the optimal target and corresponding profit from neighbor nodes;

[0076] 3.4) Forming a local bidding set: Integrating its own and its neighbors' bidding information to form a local "node-goal-benefit" set, identifying the set of neighbors with the same goal selection as itself, and forming a conflict set on the current goal;

[0077] 3.5) Determine if there is a target conflict: If the conflict set is empty, then the current target uniquely corresponds to the current UAV, which is considered to be without conflict. Directly confirm the current allocation result and go to step 3.8); If there is a conflict, go to step 3.6.

[0078] 3.6) Compare its own and its neighbors' earnings to determine if it has the highest bid: The drone compares its current earnings with the earnings of other neighbors in the conflict set. Based on the earnings comparison results, it determines whether it is the highest earner. If its own earnings are the highest in the conflict set, it retains the current bidding plan and proceeds to step 3.8); otherwise, it proceeds to step 3.7) to adjust the bid.

[0079] 3.7) Update the target price and recalculate the target selection: If the bid at this node is not advantageous, adjust the target price and update the current target selection and corresponding profit based on the new price, and return to step 3.2) to restart a new round of bidding.

[0080] 3.8) Output the final position allocation result: When the effective UAV completes the position selection and there is no conflict, the UAV confirms the current allocation result and completes the local position allocation (that is, completes the position allocation of the UAV).

[0081] All effective UAVs are assigned positions using steps 3.1) to 3.8), and the overall position assignment results are obtained by combining the position assignment results of each effective UAV.

[0082] 4) Recording and evaluating the effectiveness of global array position allocation results: Output the global array position allocation matrix and calculate the overall allocation efficiency index as a metric for evaluating the allocation results;

[0083] To describe the global array position allocation scheme at any given time, a binary decision variable a is introduced. ij The final global matrix A is:

[0084]

[0085] Assuming a virtual central node exists, the position allocation results of each node (i.e., each effective UAV) are uniformly collected and evaluated posteriorly. This evaluation not only outputs the global position allocation relationship but also, based on the constructed comprehensive performance function,... Calculate the overall average benefit index J assign As a quantitative evaluation standard for task matching quality, J assign The specific definition is:

[0086]

[0087] In the formula, For drones v i Assigned to target position t j The overall performance value is shown in formula (5), J assign It reflects the average level of efficiency across all current distribution relationships.

[0088] Example:

[0089] like Figure 1 The aircraft position reconstruction method based on the local conflict resolution auction method in this embodiment includes:

[0090] S1: Multi-UAV System Modeling and Position Allocation Matrix Definition:

[0091] A multi-UAV formation positioning model is established, defining the UAV state variables and target positions. In this embodiment, the formation consists of 5 UAVs, with a preset formation configuration as follows: Figure 2 As shown, the positions are arranged in a certain spatial formation according to mission requirements. Specifically, the formation expects 5 positions T = {t1, ..., t5}, and the expected positions are p. t1 =[15,48,150]',p t2 =[43,19,150]',p t3 =[-13,19,150]',p t4 =[43,76,150]' and p t5 =[-13,76,150]'; Define a drone formation of 5 drones V = {v1,...,v5}, with the drones positioned as pv1 =[40,-40,50]',p v2 =[-120,-25,70]',p v3 =[170,-35,40]',p v4 =[-20,-30,50]' and p v5 =[100,15,60]', by defining the position allocation matrix, the feasible matching relationship between each UAV and each candidate position is modeled, and allocation constraints are introduced. The initial allocation matrix A defaults to:

[0092]

[0093] Among them, if the array position t1 is assigned to the UAV v1, denoted as v1~t1, then the remaining allocation relationship can be expressed as v2~t2, v3~t3, v4~t4, v5~t5;

[0094] S2: Construction of the indicator performance function:

[0095] like Figure 3 As shown, to measure the merits of assigning different UAVs to different positions, a comprehensive performance function integrating three types of indicators is constructed, including relative distance, heading consistency, and position importance. These indicators reflect the spatial proximity between the UAV's current position and the target position, assess the consistency between the UAV's heading and the direction of the line connecting the target position, and assign weights to positions according to mission requirements to ensure priority allocation of key positions. The details are as follows:

[0096] Range performance metrics: To reflect the effective proximity between the UAV and the target array, the normalized range performance is defined as follows:

[0097]

[0098] In the formula, r ij Indicates drone v i With target position t j The relative distance between them; The ratio ||r| represents the average distance between all effective UAVs and the target array. ij ‖ / δ r The distances under different mission scenarios were normalized to ensure the scale invariance of the performance function; the exponential form compresses the distance performance value within the interval (0,1), so that the smaller the distance, the less maneuvering is required for the UAV to reach the target position, and the higher the performance.

[0099] Heading alignment effectiveness index: To quantify the degree of alignment between the UAV's current direction of motion and the line-of-sight direction of the target position, the heading alignment effectiveness is defined as follows:

[0100]

[0101] In the formula, the average angular deviation is denoted as Angle term defined as The smaller the angle, the closer the current course is to the target direction, the smaller the adjustment required, and the higher the course efficiency. Indicates drone v i The velocity vector; the exponential form is also used to compress heading alignment performance values ​​to a comparable range and to emphasize heading smoothness;

[0102] Position Importance Performance Index: To reflect the differences in the importance of target positions, the following exponential decay function is defined:

[0103]

[0104] in, The attenuation coefficient is adjustable. The index represents the target location; the smaller the target position number, the higher its importance in the formation, and therefore the higher its priority; in this embodiment, and

[0105] Comprehensive performance function: By weighting and fusing these three types of indicators, a comprehensive performance value matrix for UAV-position matching is obtained, which serves as the basis for auction decisions;

[0106]

[0107] In this embodiment, and

[0108] S3: Array position allocation process based on local conflict resolution auction:

[0109] The algorithm for position allocation based on local conflict resolution auction is designed, enabling each UAV to independently calculate its optimal target based on its own performance evaluation results. In this step, each UAV independently selects the optimal position based on its own performance evaluation results and submits a bid in the form of an offer. When multiple UAVs select the same position, the local conflict resolution mechanism is activated. The specific process is as follows: Figure 4 As shown, conflicting drones raise their bids for the position, creating upward price pressure. Drones with insufficient bids are forced to abandon the competition and move to the second-best position. This process is repeated until all position conflicts are eliminated and the allocation is stable.

[0110] To illustrate the execution process of the auction algorithm in task allocation, a partial round of a typical scheduling process is analyzed: In the first round of the auction, the five drones need to independently make allocation decisions based on the current mission objective effectiveness. Each drone calculates its comprehensive effectiveness value based on its position, orientation, and the importance of each target, and then selects the most advantageous target as its bidding object; the initial allocation phase is as follows: Figure 5 As shown, due to the differences in the task effectiveness evaluation results, all five drones selected target 1 as the preferred target without price interference. At this time, target 1 became the most competitive position, and all drones calculated it as the bidder with the highest benefit. Its maximum effective benefits were 0.8218, 0.7489, 0.6517, 0.8312 and 0.8145 respectively. Since target 1 received multiple bids at the same time, the system entered the conflict resolution stage.

[0111] In subsequent rounds, the bids for drone number 3 changed most frequently; therefore, the auction process will be described from the perspective of drone number 3: the second round of the auction... Figure 6 As shown, during the conflict detection process, UAV 3's benefit was relatively low, so its bid was considered a failure. To re-compete for target allocation, UAV 3 adjusted its bid for target 1 based on the current price status, raising its price to 0.6452, which reduced the net benefit for it. Subsequently, the benefits of all available targets were recalculated, and target 2 was found to be the option with the highest current benefit of 0.5696, thus partially mitigating the conflict.

[0112] At the beginning of the third round of auction, the specific results of the five drones' positioning were as follows: Drone 1 chose Target 2, Drone 2 chose Target 3, Drone 3 chose Target 2, Drone 4 chose Target 1, and Drone 5 chose Target 2, with corresponding profits of 0.7277, 0.5931, 0.5696, 0.8378, and 0.7095 respectively. At this point, Target 2 faced a new bidding conflict, and Drone 3's bid was not the highest. Drone 3 then proactively increased its bid for Target 2 to 0.5726, forcing its profit to decrease, thus readjusting its selection. After this round of adjustment, its assessment found that the profit value of Target 4 had increased to the new highest, at 0.3897. The results of the second round of conflict resolution are as follows. Figure 7 As shown;

[0113] In the fourth round of conflict resolution, Drone 3 further attempted to bid on Target 4, but was forced to update its bid again due to its unfavorable cost-effectiveness. Drone 3 raised its bid for Target 4 to 0.3912, diverting some competition to Target 5. At this point, Target 5's optimal cost-effectiveness was 0.2659. The drones had formed a relatively stable position selection: Drone 1 continued to choose Target 2, Drone 2 chose Target 3, Drone 3 ultimately chose Target 5, Drone 4 persisted with Target 1, and Drone 5 remained on Target 4. The conflict was thus completely resolved. Figure 8 As shown, with the elimination of conflicts and the dynamic adjustment of the price mechanism, the drones gradually converge, eventually achieving a set of conflict-free and stable position allocation results.

[0114] Through this mechanism of local conflict detection and autonomous update, the system continuously iterates and advances, enabling all participating UAVs to complete the collaborative allocation of target positions without the need for central coordination. In each round, only individuals with conflicts will recalculate target selection. This local response mechanism improves the efficiency and robustness of the algorithm. Ultimately, the system will gradually converge to a conflict-free and stable task position configuration.

[0115] The above-mentioned array allocation process embodies the characteristics of distributed, autonomous decision-making, and collaborative coordination. It can effectively adapt to the dynamic changes in the system caused by faults, communication interruptions, or target changes during the mission. It is one of the key foundations for building a highly reliable, multi-agent collaborative system.

[0116] S4: Recording and Evaluation of Global Array Position Allocation Results:

[0117] Suppose there exists a virtual central node used to collect global array position allocation results and perform posterior evaluation. This node not only outputs the final UAV-array position matching relationship, but also calculates the overall average benefit index based on the aforementioned comprehensive efficiency function.

[0118] The final stable state yields of 0.7282, 0.5936, 0.2659, 0.8377 and 0.5556 for each UAV, respectively. By calculating the overall allocation efficiency formula, it can be seen that the efficiency value fluctuates greatly in the early stage of the auction process, and then quickly converges to a stable level. This indicates that the present invention achieves convergence in a small number of iterations and has the ability to quickly resolve conflicts and achieve global stability in multi-conflict scenarios.

[0119] Depend on J can be obtained assign=0.9752, reflecting the average efficiency level of all current allocation relationships. The value of this indicator ranges from 0 to 1. The closer it is to 1, the better the current position allocation result matches the state of each UAV, and the better the overall formation operation. Conversely, a lower value indicates that there is a certain deviation between the allocation result and the current state of the UAVs, which may lead to a decrease in formation operation efficiency. The change of this indicator is a relative quantity, so it can not only reflect the dynamic changes in the formation operation process, but also show sudden changes or significant fluctuations when there are abnormal situations such as external environmental disturbances or aircraft failures, thus providing a direct basis for anomaly detection. In addition, during the formation execution, continuously recording and analyzing the change curve of this indicator can be used to evaluate the effectiveness of the position reconstruction strategy. For example, if the indicator can recover and remain at a high level in a short time after a node failure, it indicates that the reconstruction strategy has strong adaptability and robustness.

[0120] In summary, the aircraft position reconfiguration method based on local conflict resolution auction of the present invention can efficiently complete the formation position reconfiguration task in complex dynamic environments, significantly improving the system's task completion rate and robustness.

Claims

1. A method for reconstructing aircraft positions based on the local conflict resolution auction method, characterized in that, include: 1) Modeling of multi-UAV systems, defining the array position allocation matrix, objective function, and allocation constraints; 2) Construct the indicator performance function; 3) Design a position allocation algorithm based on local conflict resolution auction, so that each UAV can independently calculate its optimal target based on its own performance evaluation results; 4) Recording and evaluating the effectiveness of global array position allocation results.

2. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 1, characterized in that, In step 1), the allocation constraints include: ① Each valid UAV must be assigned to a target position; ② Each target position can only be assigned to one UAV. The effective set of dynamic changes of UAVs is denoted as V = {v1, v2, ..., v...} N }, v i The i-th drone that is still capable of performing the mission at the current moment is represented by N, where N is the number of active drones at the current moment. This indicates that the effective drone v i Assigned to target position t j The efficiency value; a ij ∈[0,1] represents the corresponding binary decision variables; the target matrix set is defined as T={t1,…,t M }, N≤M; the matrix for assigning positions is defined as A=[a ij ] N×M The objective function reflects the expected effect of the overall array allocation based on the contributions of each individual UAV.

3. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 1, characterized in that, In step 2), a comprehensive performance function integrating three types of indicators is constructed, including relative distance, heading consistency, and position importance.

4. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 3, characterized in that, Step 2) includes: defining the normalized distance performance: In the formula, r ij Indicates drone v i With target position t j The distance between them; This represents the average distance between all drones and the target array. Define heading alignment effectiveness: In the formula, the average angular deviation is denoted as Angle term defined as Indicates drone v i The velocity vector; The importance and effectiveness of the formation positions are as follows: in, The attenuation coefficient is adjustable. Indicates the index of the target array position; Overall performance function: In the formula, the weighting coefficients satisfy the constraints. and N represents the number of active UAVs at the current moment; j = 1, 2, ..., M, where M is the number of target positions.

5. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 1, characterized in that, In step 3), inspired by the consensus auction algorithm, a position allocation algorithm based on local conflict resolution auction is designed. In this algorithm, each valid UAV locally calculates its effectiveness for each target position, obtains the comprehensive effectiveness vector of the UAV, and infers a globally conflict-free position allocation result through an iterative bidding process.

6. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 5, characterized in that, In step 3), the array position allocation algorithm based on local conflict resolution auction includes: Initial target selection: Each UAV will select its current optimal target based on its overall performance vector. In the formula, and γ j These represent the current assignment of the drone to the target position t. j The overall efficiency, local price, and optimal net profit, therefore, j * Corresponding benefits This represents the current maximum net profit; Conflict resolution mechanism: The core of the distributed auction algorithm lies in resolving potential conflicts through a revenue-driven price adjustment mechanism; each drone does not need to publicly disclose its complete performance matrix, but only needs to broadcast its current bid pair (j) to its neighbors. * ,γ * When multiple neighboring drones select the same target position, a conflict is considered to exist, and the conflict set is defined as follows: in For the target position selection of neighbor m; the drone will achieve its optimal benefit γ * Maximum benefit of other nodes in conflict concentration Compare, if γ * <γ max Then the drone updates the price of the current target. In the formula, k is the discrete iteration number of the current auction algorithm. For target j * The quote after the k-th update and the quote after the (k+1)-th update. It is a small positive number used to ensure the convergence of the algorithm; if there is no conflict, or the drone has the highest reward in the conflict set, it confirms the current allocation result.

7. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 5, characterized in that, Step 3) includes: 3.1) Calculate the target effectiveness index: Based on the current state information, each effective UAV calculates its effectiveness value for all target positions. The effectiveness function integrates relative distance, heading angle and the importance of target positions to obtain the comprehensive effectiveness vector of the UAV. 3.2) Select the optimal target and update the bid based on local benefits: Each effective UAV selects the target with the maximum net profit as the current bidding target based on the difference between its effectiveness and the current price, and records the corresponding profit value to achieve local optimal allocation; 3.3) Broadcast its own bid and receive neighbor bid information: Broadcast its current bidding target and corresponding profit to neighbor nodes through the communication link, and receive the optimal target and corresponding profit from neighbor nodes; 3.4) Forming a local bidding set: Integrating its own and its neighbors' bidding information to form a local "node-goal-benefit" set, identifying the set of neighbors with the same goal selection as itself, and forming a conflict set on the current goal; 3.5) Determine if there is a target conflict: If the conflict set is empty, then the current target uniquely corresponds to the current UAV, which is considered to be without conflict. Directly confirm the current allocation result and go to step 3.8); If there is a conflict, go to step 3.

6. 3.6) Compare its own and its neighbors' earnings to determine if it has the highest bid: The drone compares its current earnings with the earnings of other neighbors in the conflict set. Based on the earnings comparison results, it determines whether it is the highest earner. If its own earnings are the highest in the conflict set, it retains the current bidding plan and proceeds to step 3.8); otherwise, it proceeds to step 3.7) to adjust the bid. 3.7) Update the target price and recalculate the target selection: If the bid at this node is not advantageous, adjust the target price and update the current target selection and corresponding profit based on the new price, and return to step 3.2) to restart a new round of bidding. 3.8) Output the final position allocation result: When a valid UAV completes position selection and there are no conflicts, the UAV confirms the current allocation result and completes the position allocation for that UAV.

8. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 5, characterized in that, In step 4), the global array position allocation result is output by integrating the array position allocation results of each UAV; the overall allocation efficiency index is calculated as a metric for evaluating the allocation result.

9. The aircraft position reconstruction method based on the local conflict resolution auction method as described in claim 5, characterized in that, Step 4) includes: introducing a binary decision variable a to describe the global array position allocation scheme at any given time. ij The final global matrix A is: Assuming a virtual central node exists, it uniformly collects and performs posterior evaluation of the position allocation results for each UAV. This evaluation not only outputs the global position allocation matrix but also, based on the constructed comprehensive performance function,... Calculate the overall average benefit index J assign As a quantitative evaluation standard for task matching quality, J assign The specific definition is: in, For drones v i Assigned to target position t j The overall performance value, J assign It reflects the average level of efficiency across all current distribution relationships.

Images