A method and device for establishing a high-speed aircraft cluster information fusion architecture

Abstract

Embodiments of the present invention provide a method and apparatus for establishing an information fusion architecture for a high-speed aircraft cluster. The method includes: dividing aircraft nodes within a high-speed aircraft cluster into perception clusters based on multiple targets to be perceived, thereby obtaining perception clusters corresponding to each target to be perceived; for any perception cluster, selecting a perception cluster head from the aircraft nodes within the perception cluster, and dividing aircraft nodes within the perception cluster other than the perception cluster head into information fusion clusters, thereby obtaining multiple information fusion clusters; for any information fusion cluster, selecting an information fusion cluster head from the aircraft nodes within the information fusion cluster; and establishing a three-layer information fusion architecture consisting of a joint perception cluster head, information fusion cluster head, and cluster member nodes based on the perception cluster head, each information fusion cluster head, and aircraft nodes within each information fusion cluster other than the information fusion cluster head, i.e., cluster member nodes, thereby reducing the average latency and computational cost of system positioning information fusion.

Description

Technical Field

[0001] The present invention relates to the technical field of high-speed aircraft clusters, and in particular to a method and device for establishing a high-speed aircraft cluster information fusion architecture. Background Art

[0002] Utilizing high-speed aircraft swarms to collaboratively perceive targets in the environment is a crucial task in adversarial scenarios. Currently, high-speed aircraft swarms have become an emerging force in these scenarios. Collaborative localization is a crucial component of collaborative perception and a key topic in the fields of navigation and signal processing. Information fusion within a high-speed aircraft swarm for collaborative localization involves utilizing multiple high-speed aircraft, or aircraft nodes, to perceive and locate potential non-cooperative targets in a three-dimensional environment. These nodes may be equipped with different types of sensors or located in different geographic locations, resulting in heterogeneous information acquisition. Fusion of this localization information requires preliminary filtering or error compensation. Due to differences in computing and communication resources between aircraft nodes, aircraft nodes with abundant communication and computing resources can undertake more filtering work, reducing overall system latency. Collaboration within the swarm allows for the sharing and fusion of perception data, improving positioning accuracy.

[0003] Existing research categorizes aircraft cluster information fusion architectures into three types: centralized, hierarchical, and decentralized, based on the data fusion algorithm used. Centralized architectures employ a single information fusion unit, which communicates with all aircraft and is responsible for exchanging and processing positioning information with all aircraft within the cluster. Centralized architectures utilize the positioning information of all nodes during positioning, effectively reducing errors. However, centralized architectures place high demands on the communication capabilities of the information fusion unit, making them difficult to directly apply in real-world scenarios. Furthermore, centralized architectures lack adaptability to emergencies. If the information fusion unit fails, positioning data from all aircraft cannot be processed in a timely manner, resulting in a loss of positioning capability. Hierarchical architectures still employ a single information fusion master unit. Unlike centralized architectures, however, aircraft perform local processing before transmitting data to the fusion unit. The actual data transmitted is feature-level data, which reduces communication overhead to a certain extent. However, the problem of system loss of positioning capability in emergencies still exists. Decentralized architectures lack a separate information fusion unit, providing equal status for all aircraft. Aircraft can communicate and exchange information with each other, offloading computational demands to neighboring nodes. Due to the absence of information fusion units, the distributed architecture is more robust to a certain extent. However, each aircraft node in the distributed architecture has a communication link with its neighboring nodes, which increases the pressure of communication and information fusion to a certain extent. In addition, the position of the aircraft changes rapidly in high-speed scenarios, and the established communication links have a high risk of breaking. Therefore, it is not suitable for high-speed aircraft collaborative positioning scenarios.

[0004] In view of this, in response to the challenges faced by the above architecture and the needs of high-speed aircraft collaborative positioning scenarios, it is necessary to propose a customized high-speed aircraft cluster information fusion architecture to reduce the average delay and computational consumption of system positioning information fusion. Summary of the Invention

[0005] In a first aspect, an embodiment of the present invention provides a method for establishing a high-speed aircraft cluster information fusion architecture, the method comprising:

[0006] According to multiple targets to be sensed, the aircraft nodes in the high-speed aircraft cluster are divided into perception clusters to obtain the perception clusters corresponding to the targets to be sensed;

[0007] For any perception cluster, a perception cluster head is selected from the aircraft nodes in the perception cluster, and the aircraft nodes in the perception cluster other than the perception cluster head are divided into information fusion clusters to obtain multiple information fusion clusters; for any information fusion cluster, an information fusion cluster head is selected from the aircraft nodes in the information fusion cluster; based on the perception cluster head, each information fusion cluster head, and the aircraft nodes in each information fusion cluster other than the information fusion cluster head, i.e., cluster member nodes, a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is established; wherein, the perception cluster head is located at the top layer, the information fusion cluster head is located at the middle layer, and the cluster member nodes are located at the bottom layer, the perception cluster head is communicated with each information fusion cluster head, and each information fusion cluster head is communicated with, and the information fusion cluster head is communicated with the cluster member nodes in the information fusion cluster to which it belongs.

[0008] In some implementations of the first aspect, aircraft nodes in a high-speed aircraft cluster are divided into perception clusters according to multiple targets to be perceived, to obtain perception clusters corresponding to each target to be perceived, including:

[0009] Calculate the comprehensive feature similarity between any two aircraft nodes in the high-speed aircraft cluster based on the speed similarity and distance similarity between any two aircraft nodes in the high-speed aircraft cluster;

[0010] According to the comprehensive feature similarity between any two aircraft nodes in the high-speed aircraft cluster, the aircraft nodes in the high-speed aircraft cluster are clustered to obtain multiple clusters;

[0011] Based on the communication link relevance and perception capability of the sub-clusters, a model objective function is constructed. The number and diameter constraints of the aircraft nodes in the sub-clusters are used as model constraints. Based on the model objective function and model constraints, a secondary clustering optimization model is constructed.

[0012] By solving the secondary clustering optimization model, each cluster is divided into multiple subclusters, and the subclusters are used as perception clusters;

[0013] The target perception problem is transformed into a perception cluster-to-target matching problem, and a generalized assignment model is constructed based on this problem. The generalized assignment model is then solved to match multiple perception clusters with multiple targets to be perceived, and obtain the perception cluster corresponding to each target to be perceived.

[0014] In some implementations of the first aspect, aircraft nodes other than the sensing cluster head within the sensing cluster are divided into information fusion clusters to obtain multiple information fusion clusters, including:

[0015] The silhouette coefficient algorithm is used to determine the number of information fusion clusters corresponding to the perception cluster;

[0016] Combined with the number of information fusion clusters corresponding to the perception cluster, the spectral clustering algorithm is used to cluster the aircraft nodes in the perception cluster except the perception cluster head according to the attribute data of the aircraft nodes in the perception cluster except the perception cluster head, and multiple information fusion clusters are obtained, and their number is consistent with the number of information fusion clusters.

[0017] In some implementations of the first aspect, selecting an information fusion cluster head from aircraft nodes within an information fusion cluster includes:

[0018] For any aircraft node in the information fusion cluster, the distance difference between it and multiple neighboring aircraft nodes in the information fusion cluster is calculated based on the approximate signal propagation algorithm in the loss attenuation space, and the distance standard deviation is calculated based on this. The advantage score of the aircraft node is calculated based on the corresponding distance standard deviation and computing power of the aircraft node.

[0019] The aircraft node with the highest advantage score is selected from the aircraft nodes in the information fusion cluster as the information fusion cluster head.

[0020] In some implementations of the first aspect, the method further includes:

[0021] For any perception cluster, each information fusion cluster under the perception cluster is regarded as an intelligent agent. Based on this, the data transmission decision process of the three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is constructed as a Markov decision model, and a multi-agent reinforcement learning algorithm is used to solve the Markov decision model; among them, the reward function of the Markov decision model is a reward function that combines the guiding reward and the sparse reward.

[0022] In some possible implementations of the first aspect, the Ad-Hoc protocol is used to achieve communication connection between the perception cluster head and each information fusion cluster head, and between each information fusion cluster head; the AP-AC communication protocol is used to achieve communication connection between the information fusion cluster head and the cluster member nodes in the information fusion cluster to which it belongs, wherein the information fusion cluster head is the AP and the cluster member nodes are the AC.

[0023] In some implementations of the first aspect, data is transmitted between the information fusion cluster head and the cluster member nodes in the information fusion cluster to which it belongs in a TDMA manner.

[0024] In a second aspect, an embodiment of the present invention provides a device for establishing a high-speed aircraft cluster information fusion architecture, the device comprising:

[0025] A division module is used to divide the aircraft nodes in the high-speed aircraft cluster into perception clusters according to multiple targets to be perceived, and obtain a perception cluster corresponding to each target to be perceived;

[0026] A module is established for selecting a perception cluster head from the aircraft nodes in any perception cluster, and dividing the aircraft nodes in the perception cluster except the perception cluster head into information fusion clusters to obtain multiple information fusion clusters; for any information fusion cluster, an information fusion cluster head is selected from the aircraft nodes in the information fusion cluster; a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is established based on the perception cluster head, each information fusion cluster head, and the aircraft nodes in each information fusion cluster except the information fusion cluster head, i.e., cluster member nodes; wherein the perception cluster head is located at the top layer, the information fusion cluster head is located at the middle layer, and the cluster member nodes are located at the bottom layer, the perception cluster head is communicated with each information fusion cluster head, and each information fusion cluster head is communicated with, and the information fusion cluster head is communicated with the cluster member nodes in the information fusion cluster to which it belongs.

[0027] In a third aspect, an embodiment of the present invention provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described above.

[0028] In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to enable a computer to execute the method described above.

[0029] In an embodiment of the present invention, a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node can be established for the high-speed aircraft collaborative positioning scenario to reduce the average delay and computing consumption of system positioning information fusion.

[0030] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The above and other features, advantages, and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are provided for a better understanding of the present invention and do not constitute a limitation of the present invention. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0032] Figure 1 A flowchart of a method for establishing a high-speed aircraft cluster information fusion architecture provided by an embodiment of the present invention;

[0033] Figure 2 A schematic diagram of a high-speed aircraft collaborative positioning scenario provided by an embodiment of the present invention;

[0034] Figure 3 A schematic diagram of a three-layer information fusion architecture of a joint sensing cluster head, information fusion cluster head, and cluster member nodes provided by an embodiment of the present invention;

[0035] Figure 4 A schematic diagram of a silhouette coefficient calculation result provided by an embodiment of the present invention;

[0036] Figure 5 A schematic diagram of information fusion cluster division results provided by an embodiment of the present invention;

[0037] Figure 6 A schematic diagram of the election results of an information fusion cluster head provided by an embodiment of the present invention;

[0038] Figure 7 A structural diagram of a device for establishing a high-speed aircraft cluster information fusion architecture provided by an embodiment of the present invention;

[0039] Figure 8 The figure is a structural diagram of an exemplary electronic device capable of implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0040] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0041] In addition, the term "and / or" in this invention merely describes an association relationship between related objects, indicating that three possible relationships exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Furthermore, the character " / " in this invention generally indicates that the related objects are in an "or" relationship.

[0042] In order to solve the technical problems arising from the background technology, the embodiments of the present invention provide a method, apparatus, device, and storage medium for establishing an information fusion architecture for a high-speed aircraft cluster. Specifically, the aircraft nodes within a high-speed aircraft cluster are divided into perception clusters based on multiple targets to be perceived, thereby obtaining perception clusters corresponding to each target to be perceived. For any perception cluster, a perception cluster head is selected from the aircraft nodes within the perception cluster, and the aircraft nodes within the perception cluster other than the perception cluster head are divided into information fusion clusters to obtain multiple information fusion clusters. For any information fusion cluster, an information fusion cluster head is selected from the aircraft nodes within the information fusion cluster. Based on the perception cluster head, each information fusion cluster head, and the aircraft nodes within each information fusion cluster other than the information fusion cluster head, i.e., cluster member nodes, a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member nodes is established to reduce the average latency and computational cost of system positioning information fusion.

[0043] In the following, with reference to the accompanying drawings, a method, apparatus, device and storage medium for establishing a high-speed aircraft cluster information fusion architecture provided by the embodiments of the present invention will be described in detail through specific embodiments.

[0044] Figure 1 A flowchart of a method for establishing a high-speed aircraft cluster information fusion architecture is provided in an embodiment of the present invention. Figure 1 As shown, the method 100 for establishing a high-speed aircraft cluster information fusion architecture may include:

[0045] S110 , dividing the aircraft nodes in the high-speed aircraft cluster into perception clusters according to the multiple targets to be perceived, and obtaining a perception cluster corresponding to each target to be perceived.

[0046] In some embodiments, the comprehensive feature similarity between any two aircraft nodes in the high-speed aircraft cluster can be calculated based on the speed similarity and distance similarity between any two aircraft nodes in the high-speed aircraft cluster. Based on the comprehensive feature similarity between any two aircraft nodes in the high-speed aircraft cluster, the aircraft nodes in the high-speed aircraft cluster are clustered to obtain multiple clusters. According to the communication link correlation and perception capability of the sub-clusters, a model objective function is constructed, and the aircraft node quantity constraint and diameter constraint for the sub-clusters are used as model constraints. According to the model objective function and model constraints, a secondary clustering optimization model is constructed. By solving the secondary clustering optimization model, each cluster is divided into multiple sub-clusters, and the sub-clusters are used as perception clusters. The target perception problem is converted into a perception cluster-to-target matching problem, thereby constructing a generalized assignment model, and solving the generalized assignment model to match multiple perception clusters with multiple targets to be perceived, and obtain perception clusters corresponding to each target to be perceived.

[0047] As an example, Figure 2 As shown, there are N aircraft nodes in a certain perception cluster as positioning anchor nodes, and the set The three-dimensional space coordinates of the aircraft node n are represented as p n , N aircraft nodes locate the same target m to be sensed. The target to be sensed here is a non-cooperative target, which is the target to be sensed and positioned by our high-speed aircraft cluster. Its position is uncertain for us, but we have a priori approximate position of it, which is expressed as q m .

[0048] S120, for any perception cluster, select a perception cluster head from the aircraft nodes in the perception cluster, and divide the aircraft nodes in the perception cluster except the perception cluster head into information fusion clusters to obtain multiple information fusion clusters; for any information fusion cluster, select an information fusion cluster head from the aircraft nodes in the information fusion cluster; based on the perception cluster head, each information fusion cluster head, and the aircraft nodes in each information fusion cluster except the information fusion cluster head, i.e., cluster member nodes, a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member nodes is established.

[0049] like Figure 3 As shown in Figure 1, in the three-layer information fusion architecture, the perception cluster head is located at the top layer, the information fusion cluster head is located in the middle layer, and the cluster member nodes are located at the bottom layer. The perception cluster head is connected to each information fusion cluster head, and each information fusion cluster head is connected to each cluster member node within its information fusion cluster.

[0050] Optionally, a peer-to-peer (Ad-Hoc / IBSS) protocol is used to establish communication between the perception cluster head and each information fusion cluster head, and between each other. This ensures direct communication between cluster heads. When a certain information fusion cluster head experiences a high computational load within a short period of time, it can offload computational tasks to other information fusion cluster heads or perception cluster heads. Similarly, the perception cluster head can offload information fusion tasks to the information fusion cluster head to achieve collaboration and reduce the total duration of a single information fusion. The AP-AC communication protocol is used to establish communication between the information fusion cluster head and the cluster member nodes within its information fusion cluster. The information fusion cluster head operates in wireless access point mode (AP), and the cluster member nodes operate in client mode (AC / Managed). Since the AP-AC communication protocol only requires a communication link between the cluster member nodes and the information fusion cluster head, and does not require direct connections between cluster member nodes, it improves the stability of the fusion architecture to a certain extent.

[0051] It should be noted that the sensing cluster head, as the cluster head in the traditional sense, is responsible for overall planning within the sensing cluster, mastering the status and information of the aircraft nodes within the sensing cluster, allocating channel resources, and inter-cluster communication, etc. The information fusion cluster head is responsible for data transmission, relaying, computing, etc., and provides load balancing capabilities for the system to a certain extent. The cluster member nodes are responsible for data transmission with the information fusion cluster head in a Time Division Multiple Access (TDMA) manner to achieve information fusion.

[0052] In some embodiments, the above-mentioned division of the information fusion clusters for the aircraft nodes other than the sensing cluster head within the sensing cluster to obtain multiple information fusion clusters may include:

[0053] Use the silhouette coefficient algorithm to determine the number of information fusion clusters corresponding to the sensing cluster, and then combine the number of information fusion clusters corresponding to the sensing cluster. Use the spectral clustering algorithm to cluster the aircraft nodes other than the sensing cluster head within the sensing cluster according to the attribute data (such as position, speed, etc.) of the aircraft nodes other than the sensing cluster head within the sensing cluster to obtain multiple information fusion clusters, and the number thereof is consistent with the number of information fusion clusters.

[0054] Among them, the silhouette coefficient algorithm can be as shown in the following formula:

[0055] <00​​​​​​​​​​​​After determining the number of information fusion clusters, it is necessary to further determine the partitioning algorithm to complete the partitioning of the information fusion clusters. The state of the high-speed aircraft cluster is dynamically changing, and the relationship, position, speed and other information between nodes will change over time. Due to its graph-based nature, spectral clustering can flexibly adapt to this dynamic change and re-cluster according to the similarity of real-time changes. Therefore, the spectral clustering algorithm is used here to construct information fusion clusters. The idea of ​​spectral clustering is derived from spectral graph theory, which regards all data as points in space and links them through an undirected weighted graph. The edge weight between two points that are farther apart is lower, and the edge weight between two points that are closer is higher. By cutting the undirected weighted graph, the sum of the weights between different subgraphs after cutting is made as low as possible to achieve the purpose of clustering. Taking position similarity as the objective function of spectral clustering, the information fusion clustering partitioning result based on spectral clustering can be as follows: Figure 5 As shown, aircraft nodes with the same color belong to the same cluster.

[0059] In some embodiments, the above-mentioned selection of the information fusion cluster head from the aircraft nodes in the information fusion cluster may include:

[0060] For any aircraft node within an information fusion cluster, the distance difference between it and its multiple neighboring aircraft nodes in the cluster is calculated using an approximate signal propagation algorithm in loss-attenuation space. This distance standard deviation is then used to calculate the dominance score of the aircraft node. This is then combined with the corresponding distance standard deviation and computing power to calculate the dominance score of the aircraft node. The aircraft node with the highest dominance score is selected from the aircraft nodes within the information fusion cluster as the cluster head.

[0061] As an example, since the information fusion cluster head undertakes part of the data transmission and information fusion tasks, it is necessary to comprehensively consider the location and computing power when electing the information fusion cluster head to reduce communication consumption and ensure low computing latency. The relative distance between aircraft nodes changes rapidly over time. Considering the power received and transmitted by the aircraft node, the distance difference between the aircraft node and the neighboring node is calculated based on the approximate signal propagation algorithm in the loss attenuation space. Specifically, let aircraft node i be any node within the next hop communication range of aircraft node j, that is, the neighboring node. Then the distance difference d between aircraft node j and aircraft node i is ji for:

[0062]

[0063] Where, P t Indicates the signal transmission power of the signal sending node, G t represents the transmitting antenna gain of the signal sending node, h t Indicates the height of the transmitting antenna of the signal sending node, h r Indicates the height of the receiving antenna of the signal receiving node, Pr It is worth noting that if aircraft node j is a signal sending node, then aircraft node i is a signal receiving node, and vice versa.

[0064] On this basis, the average distance difference between aircraft node j and its K neighbor aircraft nodes i in the information fusion cluster is calculated. and the distance standard deviation δ j , respectively:

[0065]

[0066] After this, with f j Represents the computing power of aircraft node j, which is obtained by weighted summation δ j 、f j Calculate the advantage score of aircraft node j, and based on this, elect the aircraft node with the highest advantage score in the information fusion cluster as the information fusion cluster head. Taking the perception cluster composed of 30 aircraft nodes as an example, the election result of the information fusion cluster head is as follows: Figure 6 shown.

[0067] In some embodiments, after S120 , the high-speed aircraft cluster information fusion architecture establishment method 100 may further include:

[0068] For any perception cluster, each information fusion cluster under the perception cluster is regarded as an intelligent agent. Based on this, the data transmission decision process of the three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is constructed as a Markov decision model, and a multi-agent reinforcement learning algorithm is used to solve the Markov decision model; among them, the reward function of the Markov decision model is a reward function that combines the guiding reward and the sparse reward.

[0069] As an example, the data transmission decision of the three-layer information fusion architecture is an NP-hard problem. To solve this problem, a Markov decision model is constructed here, and an algorithm based on multi-agent reinforcement learning is proposed to solve it. Each information fusion cluster is regarded as an agent, and an agent set is constructed. The time slot when a cluster member node transmits data to the information fusion cluster head is regarded as a time step.

[0070] 1. State Space

[0071] State Space It can be shown as the following formula, including the computing power of all cluster member nodes in the information fusion cluster Transmission power The amount of data perceived The computing power of the sensing cluster head f p, data volume d p .

[0072]

[0073] 2. Action Space

[0074] Action Space It can be shown as the following formula, including the data transmission α from all cluster member nodes in the information fusion cluster to the information fusion cluster head i , data transmission between information fusion cluster heads β i and data transmission from information fusion cluster head to perception cluster head γ i . represents the data transmission decision of cluster member node n in information fusion cluster j in the i-th time slot, Indicates that no transmission has been made. Indicates transmission, represents the data transmission decision of the information fusion cluster head in information fusion cluster head j to other information fusion cluster heads in the i-th time slot, It indicates that the information fusion cluster head in this cluster transmits data to the information fusion cluster head in information fusion cluster 1. represents the data transmission decision of the information fusion cluster head in information fusion cluster j to the sensing cluster head in the i-th time slot, Indicates that no transmission has been made. Indicates transmission.

[0075]

[0076] 3. Reward Function

[0077] The goal of data transmission decision-making is to minimize the total information fusion delay. This optimization objective is sparse as a reward function and is difficult to directly guide training. Therefore, a reward function that combines guided rewards with sparse rewards is proposed here, as shown below:

[0078]

[0079] Where, represents a one-time reward for the completion of the information fusion process, It represents the guiding reward of each time slot during the information fusion process, which is used to reduce the difficulty of training. It represents the ratio of the amount of data processed by information fusion cluster j to the time consumed in the i-th time slot. The average value of each time slot is reduced by training. To reduce latency.

[0080] Based on the above Markov decision process, a data transmission decision algorithm based on multi-agent reinforcement learning is proposed as follows:

[0081]

[0082]

[0083]

[0084] In summary, according to the embodiments of the present invention, at least the following technical effects are achieved:

[0085] 1. The present invention aims at the collaborative perception scenario of high-speed aircraft and rationally establishes communication links and information fusion architecture. In response to the challenges faced by information fusion architecture in existing research, a hierarchical information fusion structure is proposed, which adopts different communication link methods between different nodes, adopts AP-AC communication protocol within the group, and adopts Ad-Hoc protocol between cluster heads, thereby improving the stability of the communication link. The concept of information fusion cluster head is proposed to reduce the computing resource consumption of the perception cluster head. The data transmission within different information fusion clusters can be executed in parallel, and load balancing is achieved to a certain extent. In addition, in the information fusion architecture of the present invention, even if a node cannot continue to work, the system still has positioning and information fusion capabilities.

[0086] 2. A multi-agent reinforcement learning algorithm was designed to address the data transmission issues between nodes in the information fusion architecture. Each information fusion cluster was configured as an agent, avoiding the large action dimension and low training efficiency issues of traditional centralized reinforcement learning. The reward function design considered the impact of reward sparsity on training effectiveness, combining guided rewards with sparse rewards to reduce training difficulty.

[0087] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0088] The above is an introduction to a method embodiment. The following further illustrates the solution of the present invention through an apparatus embodiment.

[0089] Figure 7 A structural diagram of a high-speed aircraft cluster information fusion architecture establishment device provided by an embodiment of the present invention, such as Figure 7 As shown, the high-speed aircraft cluster information fusion architecture establishment device 700 may include:

[0090] The division module 710 is used to divide the aircraft nodes in the high-speed aircraft cluster into perception clusters according to multiple targets to be perceived, and obtain a perception cluster corresponding to each target to be perceived.

[0091] Establish module 720, which is used to select a perception cluster head from the aircraft nodes in any perception cluster, and divide the aircraft nodes in the perception cluster except the perception cluster head into information fusion clusters to obtain multiple information fusion clusters; for any information fusion cluster, select an information fusion cluster head from the aircraft nodes in the information fusion cluster; establish a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node based on the perception cluster head, each information fusion cluster head, and the aircraft nodes in each information fusion cluster except the information fusion cluster head, i.e., cluster member nodes; wherein the perception cluster head is located at the top layer, the information fusion cluster head is located at the middle layer, and the cluster member nodes are located at the bottom layer, the perception cluster head is communicated with each information fusion cluster head, and each information fusion cluster head is communicated with, and the information fusion cluster head is communicated with the cluster member nodes in the information fusion cluster to which it belongs.

[0092] It is understandable that Figure 7 Each module / unit in the high-speed aircraft cluster information fusion architecture establishment device 700 has the function of realizing Figure 1 The functions of the various steps in the high-speed aircraft cluster information fusion architecture establishment method 100 shown and their ability to achieve corresponding technical effects are not described here for the sake of brevity.

[0093] Figure 8 8 is a block diagram of an exemplary electronic device capable of implementing embodiments of the present invention. Electronic device 800 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 800 may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0094] like Figure 8 As shown, the electronic device 800 may include a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the electronic device 800 may also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0095] Multiple components in the electronic device 800 are connected to the I / O interface 805, including an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0096] The computing unit 801 can be a variety of general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as method 100. For example, in some embodiments, method 100 can be implemented as a computer program product, including a computer program tangibly embodied in a computer-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the method 100 described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the method 100 in any other appropriate manner (e.g., by means of firmware).

[0097] The various embodiments described above in the present invention can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0098] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. Such program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0099] In the context of the present invention, computer-readable media can be tangible media that can contain or store a program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable media can be a computer-readable signal medium or a computer-readable storage medium. Computer-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. More specific examples of computer-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0100] It should be noted that the present invention also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute method 100 and achieve the corresponding technical effect achieved by executing the method in an embodiment of the present invention. For the sake of brevity, they will not be repeated here.

[0101] In addition, the present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the method 100 is implemented.

[0102] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present invention can be achieved. The present invention is not limited here.

[0103] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for establishing a high-speed aircraft cluster information fusion architecture, characterized in that: The method comprises: According to multiple targets to be sensed, the aircraft nodes in the high-speed aircraft cluster are divided into perception clusters to obtain the perception clusters corresponding to the targets to be sensed; For any perception cluster, a perception cluster head is selected from the aircraft nodes in the perception cluster, and the aircraft nodes in the perception cluster other than the perception cluster head are divided into information fusion clusters to obtain multiple information fusion clusters; for any information fusion cluster, an information fusion cluster head is selected from the aircraft nodes in the information fusion cluster; based on the perception cluster head, each information fusion cluster head, and the aircraft nodes in each information fusion cluster other than the information fusion cluster head, i.e., cluster member nodes, a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is established; wherein, the perception cluster head is located at the top layer, the information fusion cluster head is located at the middle layer, and the cluster member nodes are located at the bottom layer, the perception cluster head is communicated with each information fusion cluster head, and each information fusion cluster head is communicated with, and the information fusion cluster head is communicated with the cluster member nodes in the information fusion cluster to which it belongs; The information fusion cluster division is performed on the aircraft nodes except the sensing cluster head in the sensing cluster to obtain multiple information fusion clusters, including: The silhouette coefficient algorithm is used to determine the number of information fusion clusters corresponding to the perception cluster; Combined with the number of information fusion clusters corresponding to the perception cluster, the spectral clustering algorithm is used to cluster the aircraft nodes in the perception cluster except the perception cluster head according to their attribute data, and multiple information fusion clusters are obtained, the number of which is consistent with the number of information fusion clusters. The step of selecting an information fusion cluster head from the aircraft nodes in the information fusion cluster includes: For any aircraft node in the information fusion cluster, the distance difference between it and multiple neighboring aircraft nodes in the information fusion cluster is calculated based on the approximate signal propagation algorithm in the loss attenuation space, and the distance standard deviation is calculated based on this. The advantage score of the aircraft node is calculated based on the corresponding distance standard deviation and computing power of the aircraft node. The aircraft node with the highest advantage score is selected from the aircraft nodes in the information fusion cluster as the information fusion cluster head.

2. The method according to claim 1, characterized in that The method of dividing the aircraft nodes in the high-speed aircraft cluster into perception clusters according to the multiple targets to be perceived to obtain perception clusters corresponding to the respective targets to be perceived includes: Calculate the comprehensive feature similarity between any two aircraft nodes in the high-speed aircraft cluster based on the speed similarity and distance similarity between any two aircraft nodes in the high-speed aircraft cluster; According to the comprehensive feature similarity between any two aircraft nodes in the high-speed aircraft cluster, the aircraft nodes in the high-speed aircraft cluster are clustered to obtain multiple clusters; Based on the communication link relevance and perception capability of the sub-clusters, a model objective function is constructed. The number and diameter constraints of the aircraft nodes in the sub-clusters are used as model constraints. Based on the model objective function and model constraints, a secondary clustering optimization model is constructed. By solving the secondary clustering optimization model, each cluster is divided into multiple subclusters, and the subclusters are used as perception clusters; The target perception problem is transformed into a perception cluster-to-target matching problem, and a generalized assignment model is constructed based on this problem. The generalized assignment model is then solved to match multiple perception clusters with multiple targets to be perceived, and obtain the perception cluster corresponding to each target to be perceived.

3. The method according to claim 1, characterized in that The method further comprises: For any perception cluster, each information fusion cluster under the perception cluster is regarded as an intelligent agent. Based on this, the data transmission decision process of the three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is constructed as a Markov decision model, and a multi-agent reinforcement learning algorithm is used to solve the Markov decision model; among them, the reward function of the Markov decision model is a reward function that combines the guiding reward and the sparse reward.

4. The method according to claim 1, wherein The Ad-Hoc protocol is used to achieve communication between the perception cluster head and each information fusion cluster head, and between each information fusion cluster head. The AP-AC communication protocol is used to achieve communication between the information fusion cluster head and the cluster member nodes within its information fusion cluster, where the information fusion cluster head is the AP and the cluster member nodes are the AC.

5. The method according to claim 1, wherein The information fusion cluster head transmits data to the cluster member nodes in the information fusion cluster to which it belongs in a TDMA manner.

6. A device for establishing a high-speed aircraft cluster information fusion architecture, characterized in that: The device comprises: A division module is used to divide the aircraft nodes in the high-speed aircraft cluster into perception clusters according to multiple targets to be perceived, and obtain a perception cluster corresponding to each target to be perceived; A module is established for selecting a perception cluster head from the aircraft nodes in the perception cluster for any perception cluster, and performing information fusion cluster division on the aircraft nodes in the perception cluster except the perception cluster head to obtain multiple information fusion clusters; for any information fusion cluster, an information fusion cluster head is selected from the aircraft nodes in the information fusion cluster; a three-layer information fusion architecture of joint perception cluster head-information fusion cluster head-cluster member node is established based on the perception cluster head, each information fusion cluster head, and the aircraft nodes in each information fusion cluster except the information fusion cluster head, i.e., cluster member nodes; wherein the perception cluster head is located at the top layer, the information fusion cluster head is located at the middle layer, and the cluster member nodes are located at the bottom layer, the perception cluster head is in communication connection with each information fusion cluster head, and each information fusion cluster head is in communication connection, and the information fusion cluster head is in communication connection with the cluster member nodes in the information fusion cluster to which it belongs; In the establishment module, the aircraft nodes in the perception cluster except the perception cluster head are divided into information fusion clusters to obtain multiple information fusion clusters, including: The silhouette coefficient algorithm is used to determine the number of information fusion clusters corresponding to the perception cluster; Combined with the number of information fusion clusters corresponding to the perception cluster, the spectral clustering algorithm is used to cluster the aircraft nodes in the perception cluster except the perception cluster head according to their attribute data, and multiple information fusion clusters are obtained, the number of which is consistent with the number of information fusion clusters. The step of selecting an information fusion cluster head from the aircraft nodes in the information fusion cluster includes: For any aircraft node in the information fusion cluster, the distance difference between it and multiple neighboring aircraft nodes in the information fusion cluster is calculated based on the approximate signal propagation algorithm in the loss attenuation space, and the distance standard deviation is calculated based on this. The advantage score of the aircraft node is calculated based on the corresponding distance standard deviation and computing power of the aircraft node. The aircraft node with the highest advantage score is selected from the aircraft nodes in the information fusion cluster as the information fusion cluster head.

7. An electronic device, characterized in that: The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 5.

Images