A multi-view hypergraph-based matching formation method for unmanned aerial vehicles

What is AI technical title?
AI technical title is built by Patsnap AI team. It summarizes the technical point description of the patent document.
By constructing a multi-view hypergraph and fusing the relationship between the initial position of the UAV and the target position, and by using the Hungarian algorithm and the robust distributed formation control algorithm, the problem of insufficient multi-view fusion in UAV matching formation is solved, and the accurate positioning and efficient matching of UAV target positions are achieved.

CN118034382BActive Publication Date: 2026-07-14SHANXI UNIV

Patent Information

Authority / Receiving Office: CN · China
Patent Type: Patents(China)
Current Assignee / Owner: SHANXI UNIV
Filing Date: 2024-03-27
Publication Date: 2026-07-14

AI Technical Summary

Technical Problem

Existing drone matching and formation methods fail to effectively integrate the relationship between the initial position and the target position from multiple perspectives, resulting in slow optimization convergence and difficulty in obtaining a better matching relationship.

Method used

A multi-view hypergraph method is adopted to construct a hypergraph of the initial position and the target position. By normalizing the UAV coordinates and fusing the multi-view relationships, the Hungarian algorithm is used to solve the UAV matching problem. Combined with a robust distributed formation control algorithm, the UAV can fly to the target position.

Benefits of technology

It improves the accuracy and efficiency of target location matching for drones, enhances the collaborative working ability of drone swarms, and improves the adaptability and flexibility of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure CN118034382B_ABST

Patent Text Reader

Abstract

The application discloses a kind of based on multi-view hypergraph unmanned aerial vehicle matching formation method, belong to unmanned aerial vehicle intelligent control technical field.For the relationship under initial position or target position considered by most existing methods, initial position and target position relationship are not considered from the level of multi-view fusion, so it is difficult to obtain the optimal matching relationship problem.Through obtaining the initial position and target position of unmanned aerial vehicle, and normalizing position coordinates;Multi-view hypergraph is constructed, and the initial hypergraph and target hypergraph are constructed by inputting the initial position coordinates and target position coordinates of unmanned aerial vehicle;Fusion multi-view hypergraph, get fusion hypergraph relationship;Match path is obtained using hungarian algorithm;Unmanned aerial vehicle is moved to target position using unmanned aerial vehicle flight control method.The present application is based on the fusion hypergraph including initial position and target position relationship weight, can fully consider the position information of unmanned aerial vehicle, is helpful to determine target position when solving actual unmanned aerial vehicle matching formation problem.

Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) intelligent control technology, specifically relating to a UAV matching and formation method based on a multi-view hypergraph. Background Technology

[0002] Drone swarm collaboration refers to the technology of a group of drones working together to achieve a common goal or mission. Drone matching and formation technology is a key technology in this field, aiming to achieve a reasonable match between the initial position and the target position of drones by rationally allocating tasks.

[0003] Matching the initial and target positions of a drone swarm typically falls into two categories: manually assigned and computationally optimized. Manual assignment usually requires significant manpower and time and struggles to achieve optimal matching. Currently, optimization algorithms are commonly used to plan the initial and target positions of drones, i.e., computational optimization. Existing computational optimization methods can be categorized into centralized planning, distributed planning, and learning model-based planning. Centralized planning considers the overall task requirements, using mathematical modeling and algorithmic solutions to maximize overall efficiency or satisfy task-specific constraints. This method allows the system to collaboratively manage the entire drone swarm, ensuring efficient task execution throughout the space. However, centralized planning also faces challenges related to numerous optimization iterations and computational complexity. In distributed planning, each drone plans its initial and target positions based on the surrounding environment and task requirements under local collaboration. Drones exchange position information to achieve distributed matching decisions, thereby improving the system's adaptability and flexibility. However, distributed planning may also face challenges related to incomplete local information. Learning-based matching techniques utilize machine learning algorithms to learn from and train on large amounts of data, enabling the system to automatically identify and predict the relationship between the initial position and the target position of a UAV. This method can adapt to complex and dynamic environments, enabling personalized matching strategies. Common learning algorithms include deep learning and reinforcement learning, which can extract features and patterns from data, allowing the system to make better position matching decisions for the UAV. However, most methods only consider the relationship between the initial or target position, without considering the relationship between the initial and target positions from a multi-perspective fusion level. This makes it difficult to obtain a better matching relationship, further leading to slow convergence of the corresponding algorithms. Summary of the Invention

[0004] To address the issue that most existing methods only consider the relationship between the initial position or the target position, without considering the relationship between the initial position and the target position from the perspective of multi-view fusion, it is difficult to obtain a better matching relationship, which further leads to the slow convergence of the corresponding algorithm. This invention provides a UAV matching formation method based on multi-view hypergraph.

[0005] To achieve the above objectives, the present invention employs the following technical solutions:

[0006] A method for drone matching and formation based on multi-view hypergraphs includes the following steps:

[0007] Step 1: Obtain the initial position coordinates and target position coordinates of the UAV, and normalize the coordinates;

[0008] Step 1.1, Obtain the initial position coordinates of the UAV: Let Let n be the initial positions of the drones, where... This represents the initial spatial coordinates of the i-th UAV;

[0009] Step 1.2, Obtain the target location coordinates of the UAV: Let Let n be the target locations of the aforementioned n drones, where Represents the spatial coordinates of the i-th UAV target;

[0010] Step 1.3, Normalize the coordinates: Initial position coordinates The normalized formula is shown in formula (1):

[0011]

[0012] and The formula is as follows:

[0013]

[0014]

[0015] Similarly, for the target position coordinates The normalization formula is:

[0016]

[0017] Similarly, calculate and

[0018]

[0019]

[0020] Step 2: Construct a multi-view hypergraph of the UAV's position. Input the initial position coordinates and target position coordinates of the UAV to construct the initial hypergraph and the target hypergraph. A hypergraph is a graph in a general sense. Unlike a regular graph where an edge can only connect two vertices, an edge in a hypergraph can connect any number of vertices.

[0021] Step 2.1, Construct the initial hypergraph: Let G be the initial hypergraph of the UAV. s =(N,e s The hypergraph consists of N = N_t hypergraph vertices, where the vertices are the sets of drones at their initial and target positions. s ∪N t , Let e represent the set of hyperedges from the initial position viewpoint. s It contains a total of 2n superedges;

[0022] Among them, hyper-edge Represents the i-th initial position with the initial position set N. s The hyperedges constructed based on this satisfy the condition shown in formula (2):

[0023]

[0024] in This represents the distance between the i-th initial position and the p-th initial position. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down, as shown in formula (3):

[0025]

[0026] The calculation process is as follows:

[0027]

[0028] Super Edge This represents the j-th target position with the initial position set N. s The hyperedges constructed based on this satisfy the condition shown in formula (4):

[0029]

[0030] in This represents the distance between the j-th target position and the p-th initial position. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down. The calculation formula is shown in formula (5):

[0031]

[0032] The calculation process is as follows:

[0033]

[0034] The relationship between the initial position and the target position G from the initial position perspective. s (i,j) is calculated using the set of neighbors in the initial position and the set of neighbors in the initial position at the target position, as shown in formula (6).

[0035]

[0036] in, Indicates the superedge With super-edge The total number of drones included. Indicates the superedge With super-edge The number of all included drones was calculated above to obtain the relationship between the initial position and the target position from the initial position perspective.

[0037] Step 2.2, Construct the target hypergraph: Let G be the target hypergraph of the UAV. t =(N,e t The hypergraph consists of N = N_t hypergraph vertices, where the vertices are the sets of drones at their initial and target positions. s ∪N t , Let e represent the set of hyperedges from the perspective of the target location. t It contains a total of 2n superedges;

[0038] Among them, hyper-edge Represents the i-th target location with respect to the target location set N. t The hyperedges constructed based on this satisfy the condition shown in formula (7):

[0039]

[0040] in This represents the distance between the i-th target location and the p-th target location. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down, as shown in formula (8):

[0041]

[0042] The calculation process is as follows:

[0043]

[0044] Super Edge Represents the j-th initial position relative to the target position set N. t The hyperedges constructed based on this satisfy the condition shown in formula (9):

[0045]

[0046] in This represents the distance between the j-th initial position and the p-th target position. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down. The calculation formula is shown in formula (10):

[0047]

[0048] The calculation process is as follows:

[0049]

[0050] The relationship between the initial position and the target position from the perspective of the target position, G t (i,j) is calculated using the set of neighbors in the target location and the set of neighbors in the target location from the initial location, as shown in formula (11):

[0051]

[0052] in, Indicates the superedge With super-edge The total number of drones included. Indicates the superedge With super-edge The number of all included drones was used to calculate the relationship between the initial position and the target position from the target position's perspective.

[0053] Step 3: Merge the multi-view hypergraphs to obtain the fused hypergraph relations; merge the initial hypergraph and the target hypergraph to obtain the fused hypergraph G = (N, e), where e represents the set of hyperedges in the fused hypergraph, and the initial position G in the fused hypergraph is... s (v i ,v j ) and target position G t (v i ,v j The relationship weights G(v) between the two sides i ,v j The calculation is shown in formula (12):

[0054]

[0055] The calculation yields the weights of the relationship between the initial and target positions after fusion, which are then used as input for step 4.

[0056] Step 4, Obtain the matching position: To find the matching path between the initial position and the target position of the UAV, the Hungarian algorithm is applied to solve this problem. The Hungarian algorithm is a combinatorial optimization algorithm that solves the task allocation problem in polynomial time and can be used to solve the matching problem of bipartite graphs. In the UAV matching task, the bipartite graphs are the initial position and the target position of the UAV. Based on the fusion hypergraph G in Step 3, the weight relationship between the initial position and the target position is obtained and used as the input of the Hungarian algorithm. The algorithm obtains the matching with the maximum weight by solving a linear programming problem, and then outputs the matrix of the optimal matching situation, that is, the matching matrix between the initial position and the target position.

[0057] Step 5, Flying to the target location: The existing UAV flight control algorithm is used to realize the change of the UAV swarm from the initial position to the target state. This invention uses the robust distributed formation control method in the literature (Robust Distributed Formation Control of AgentsWith Higher-Order Dynamics.IEEE Control Systems Letters,2(3):495–500.) to realize the UAV flying to the target location.

[0058] Compared with the prior art, the present invention has the following advantages:

[0059] (1) The UAV matching formation method based on multi-view hypergraph proposed in this invention considers the positional relationship between the viewpoint at the initial position of the UAV and the viewpoint at the target position, and can take into account the positional information of UAV matching to the greatest extent.

[0060] (2) The UAV matching formation method based on multi-view hypergraph proposed in this invention integrates the relationship information between the initial and final viewpoints of the UAV positions in the hypergraph, which can effectively improve the accuracy of UAV target positioning. Attached Figure Description

[0061] Figure 1 This is a flowchart of a drone matching and formation method based on a multi-view hypergraph;

[0062] Figure 2 This is a step diagram of a drone matching and formation method based on a multi-view hypergraph;

[0063] Figure 3 A multi-view hypergraph-based UAV matching and formation method with different iteration numbers of UAV position maps;

[0064] Figure 4 UAV location maps based on different iterations of the Euclidean distance method. Detailed Implementation

[0065] like Figure 1 and Figure 2 As shown, a UAV matching and formation method based on multi-view hypergraphs includes the following steps:

[0066] Step 1: Obtain the initial position coordinates and target position coordinates of the UAV, and normalize the coordinates;

[0067] Step 1.1, Obtain the initial position coordinates of the UAV: Let Let n be the initial positions of the drones, where... This represents the initial spatial coordinates of the i-th UAV;

[0068] Step 1.2, Obtain the target location coordinates of the UAV: Let Let n be the target locations of the aforementioned n drones, where Represents the spatial coordinates of the i-th UAV target;

[0069] Step 1.3, Normalize the coordinates: Initial position coordinates The normalized formula is shown in formula (1):

[0070]

[0071] and The formula is as follows:

[0072]

[0073]

[0074] Similarly, for the target position coordinates The normalization formula is:

[0075]

[0076] Similarly, calculate and

[0077]

[0078]

[0079] Step 2: Construct a multi-view hypergraph of the UAV's position. Input the initial position coordinates and target position coordinates of the UAV to construct the initial hypergraph and the target hypergraph. A hypergraph is a graph in a general sense. Unlike a regular graph where an edge can only connect two vertices, an edge in a hypergraph can connect any number of vertices.

[0080] Step 2.1, Construct the initial hypergraph: Let G be the initial hypergraph of the UAV. s =(N,e s The hypergraph consists of N = N_t hypergraph vertices, where the vertices are the sets of drones at their initial and target positions. s ∪N t , Let e represent the set of hyperedges from the initial position viewpoint. s It contains a total of 2n superedges;

[0081] Among them, hyper-edge Represents the i-th initial position with the initial position set N. s The hyperedges constructed based on this satisfy the condition shown in formula (2):

[0082]

[0083] in This represents the distance between the i-th initial position and the p-th initial position. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down, as shown in formula (3):

[0084]

[0085] The calculation process is as follows:

[0086]

[0087] Super Edge This represents the j-th target position with the initial position set N. s The hyperedges constructed based on this satisfy the condition shown in formula (4):

[0088]

[0089] in This represents the distance between the j-th target position and the p-th initial position. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down. The calculation formula is shown in formula (5):

[0090]

[0091] The calculation process is as follows:

[0092]

[0093] The relationship between the initial position and the target position G from the initial position perspective. s (i,j) is calculated using the set of neighbors in the initial position and the set of neighbors in the initial position at the target position, as shown in formula (6).

[0094]

[0095] in, Indicates the superedge With super-edge The total number of drones included. Indicates the superedge With super-edge The number of all included drones was calculated above to obtain the relationship between the initial position and the target position from the initial position perspective.

[0096] Step 2.2, Construct the target hypergraph: Let G be the target hypergraph of the UAV. t =(N,e t The hypergraph consists of N = N_t hypergraph vertices, where the vertices are the sets of drones at their initial and target positions. s ∪N t , Let e represent the set of hyperedges from the perspective of the target location. t It contains a total of 2n superedges;

[0097] Among them, hyper-edge Represents the i-th target location with respect to the target location set N. t The hyperedges constructed based on this satisfy the condition shown in formula (7):

[0098]

[0099] in This represents the distance between the i-th target location and the p-th target location. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down, as shown in formula (8):

[0100]

[0101] The calculation process is as follows:

[0102]

[0103] Super Edge Represents the j-th initial position relative to the target position set N. t The hyperedges constructed based on this satisfy the condition shown in formula (9):

[0104]

[0105] in This represents the distance between the j-th initial position and the p-th target position. Indicates the superedge The number of drones included. This means taking the square root of n and rounding down. The calculation formula is shown in formula (10):

[0106]

[0107] The calculation process is as follows:

[0108]

[0109] The relationship between the initial position and the target position from the perspective of the target position, G t (i,j) is calculated using the set of neighbors in the target location and the set of neighbors in the target location from the initial location, as shown in formula (11):

[0110]

[0111] in, Indicates the superedge With super-edge The total number of drones included. Indicates the superedge With super-edge The number of all included drones was used to calculate the relationship between the initial position and the target position from the target position's perspective.

[0112] Step 3: Merge the multi-view hypergraphs to obtain the fused hypergraph relations; merge the initial hypergraph and the target hypergraph to obtain the fused hypergraph G = (N, e), where e represents the set of hyperedges in the fused hypergraph, and the initial position G in the fused hypergraph is... s (v i ,v j ) and target position G t (v i ,v j The relationship weights G(v) between the two sides i ,v j The calculation is shown in formula (12):

[0113]

[0114] The calculation yields the weights of the relationship between the initial and target positions after fusion, which are then used as input for step 4.

[0115] Step 4, Obtain the matching position: To find the matching path between the initial position and the target position of the UAV, the Hungarian algorithm is applied to solve this problem. The Hungarian algorithm is a combinatorial optimization algorithm that solves the task allocation problem in polynomial time and can be used to solve the matching problem of bipartite graphs. In the UAV matching task, the bipartite graphs are the initial position and the target position of the UAV. Based on the fusion hypergraph G in Step 3, the weight relationship between the initial position and the target position is obtained and used as the input of the Hungarian algorithm. The algorithm obtains the matching with the maximum weight by solving a linear programming problem, and then outputs the matrix of the optimal matching situation, that is, the matching matrix between the initial position and the target position.

[0116] Step 5, Flying to the target location: The existing UAV flight control algorithm is used to realize the change of the UAV swarm from the initial position to the target state. This invention uses the robust distributed formation control method in the literature (Robust Distributed Formation Control of AgentsWith Higher-Order Dynamics.IEEE Control Systems Letters,2(3):495–500.) to realize the UAV flying to the target location.

[0117] Appendix Figure 3 and attached Figure 4 As can be seen, the method achieves the same positional accuracy of the UAV after 150 iterations as the Euclidean distance method after 200 iterations, thus proving that the method can effectively improve the accuracy of UAV target positioning.

[0118] Contents not described in detail in this specification are prior art known to those skilled in the art. Although illustrative specific embodiments of the invention have been described above to facilitate understanding by those skilled in the art, it should be understood that the invention is not limited to the scope of the specific embodiments. Various modifications are readily apparent to those skilled in the art as long as they fall within the spirit and scope of the invention as defined and determined by the appended claims, and all inventions utilizing the concept of this invention are protected.

Claims

1. A method for UAV matching and formation based on multi-view hypergraphs, characterized in that: Includes the following steps: Step 1: Obtain the initial position coordinates and target position coordinates of the UAV, and normalize the coordinates; Step 1.1, Obtain the initial position coordinates of the UAV: Let for The initial position set of the drones, among which, Indicates the first The initial spatial coordinates of the drone; Step 1.2, Obtain the target location coordinates of the UAV: Let For the above A set of target locations for drones, among which, Indicates the first The spatial coordinates of the drone target; Step 2: Construct a multi-view hypergraph of the UAV's position. Input the initial position coordinates of the UAV and the target position coordinates to construct the initial hypergraph and the target hypergraph. Step 2.1, Construct the initial hypergraph: Let the initial hypergraph of the UAV be... The vertices of the hypergraph are the sets of drones at their initial and target positions. , This represents the set of hyperedges from the initial position viewpoint. Total of Strip of edge; Among them, hyper-edge Indicates the first An initial position with an initial position set The hyperedges constructed based on this satisfy the condition shown in formula (2): (2)； in Indicates the drone's first The initial position and the first The distance between the initial positions Indicates the superedge The number of drones included. Indicates to Take the square root and round down, as shown in formula (3): (3)； The calculation process is as follows: ； Super Edge Indicates the first The target locations are given by the initial location set. The hyperedges constructed based on this satisfy the condition shown in formula (4): (4)； in Indicates the first The target location and the first The distance between the initial positions Indicates the superedge The number of drones included. Indicates to The square root is taken and rounded down, and the calculation formula is shown in formula (5): (5)； The calculation process is as follows: ； The relationship between the initial position and the target position from the initial position perspective. It is calculated by combining the set of neighbors in the initial position and the set of neighbors in the initial position of the target position, as shown in formula (6). (6)； in, Indicates the superedge With super-edge The total number of drones included. Indicates the superedge With super-edge The number of all included drones was calculated above to obtain the relationship between the initial position and the target position from the initial position perspective. Step 2.2, Construct the target hypergraph: Let the target hypergraph of the UAV be... The vertices of the hypergraph are the sets of drones at their initial and target positions. , Let e represent the set of hyperedges from the perspective of the target location. t Total of Strip of edge; Among them, hyper-edge Indicates the first Target locations are set as target locations The hyperedges constructed based on this satisfy the condition shown in formula (7): (7)； in Indicates the first The target location and the first The distance between target locations Indicates the superedge The number of drones included. Indicates to Take the square root and round down, as shown in formula (8): (8)； The calculation process is as follows: ； Super Edge Indicates the first An initial position and a target position set The hyperedges constructed based on this satisfy the condition shown in formula (9): (9)； in Indicates the first The initial position and the first The distance between target locations Indicates the superedge The number of drones included. Indicates to The square root is taken and rounded down, and the calculation formula is shown in formula (10): (10)； The calculation process is as follows: ； The relationship between the initial position and the target position from the perspective of the target position, G t (i, j) is calculated using the set of neighbors in the target location and the set of neighbors in the target location from the initial location, as shown in formula (11): (11)； in, Indicates the superedge With super-edge The total number of drones included. Indicates the superedge With super-edge The number of all included drones was used to calculate the relationship between the initial position and the target position from the target position's perspective. Step 3: Fuse the multi-view hypergraphs to obtain the fused hypergraph relationships; Step 4: Use the Hungarian algorithm to obtain the matching path between the drone's initial position and the target position; Step 5, fly to the target location: Use existing UAV flight control algorithms to realize the change of the UAV swarm from the initial position to the target state.

2. The UAV matching and formation method based on a multi-view hypergraph according to claim 1, characterized in that, Step 1, which involves obtaining and standardizing the UAV's position coordinates, specifically involves the following steps: Step 1.3, Normalize the coordinates: Initial position coordinates , The normalized formula is shown in formula (1): (1)； and The formula is as follows: ，； Similarly, for the target position coordinates , The normalization formula is: ； Similarly, calculate and : ，。 3. The UAV matching and formation method based on a multi-view hypergraph according to claim 2, characterized in that, Step 3, fusing multi-view hypergraphs to obtain fused hypergraph relations, specifically involves fusing the initial hypergraph and the target hypergraph to obtain the fused hypergraph. ,in Denotes the set of hyperedges in the fused hypergraph, and the initial position in the fused hypergraph. With the target location Relationship weights The calculation is shown in formula (12): (12)； The calculation yields the weights of the relationship between the initial and target positions after fusion, which are then used as input for step 4.

4. The UAV matching and formation method based on a multi-view hypergraph according to claim 3, characterized in that, Step 4 involves using the Hungarian algorithm to obtain the matching path between the initial position and the target position of the UAV, based on the fused hypergraph from step 3. The weight relationship between the initial position and the target position is obtained and used as the input of the Hungarian algorithm. The maximum weight matching is obtained by solving a linear programming problem, and then the matrix of the optimal matching situation, that is, the matching matrix between the initial position and the target position, is output.