Unmanned aerial vehicle cluster target tracking method based on observer and hierarchical network

Abstract

The invention discloses an unmanned aerial vehicle cluster target tracking method based on an observer and a hierarchical network. The method comprises the following steps: 1) dividing unmanned aerial vehicles in an unmanned aerial vehicle cluster into three roles: a leader, a follower and an observer; n unmanned aerial vehicles form a multi-agent system, in the system, each unmanned aerial vehicle serves as an independent agent, and the multi-agent system is responsible for the whole unmanned aerial vehicle group to form an ordered formation through a leader-follower-observer model; 2) introducing a two-layer combined unmanned aerial vehicle network model, and establishing a flight ad hoc network responsible for communication in unmanned aerial vehicle cluster target tracking; and 3) after the unmanned aerial vehicle cluster realizes formation control, target tracking is carried out, and effective tracking of a single target or multiple targets is realized. According to the method, the observer concept is introduced to realize effective formation control of the cluster, so that the unmanned aerial vehicle group can more comprehensively understand the surrounding environment and make corresponding adjustment in time, and the overall maneuverability and flexibility of the unmanned aerial vehicle group are improved.

Description

Technical Field

[0001] The present invention relates to unmanned aerial vehicle control technology, and in particular to an unmanned aerial vehicle cluster target tracking method based on an observer and a hierarchical network. Background Art

[0002] The technical practices in the existing research in the field of modern drone target tracking are challenged at the application level. First, the images captured by the drone are affected by the complexity of its observation environment. Coupled with the relative motion changes between the drone and the monitored object, it is easy to cause the target to disappear in the captured image, making the tracking task difficult to continue. Secondly, the data synchronization and communication delay problems between multiple drones also have a negative impact on the collaborative efficiency and accuracy, resulting in lower adaptability, tracking efficiency, and real-time feedback capabilities when dealing with complex environments. To address these issues, a drone cluster target tracking method is needed to achieve stable tracking of fast-moving or complex background targets and significantly improve tracking efficiency. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a UAV cluster target tracking method based on observers and hierarchical networks in view of the defects in the prior art.

[0004] The technical solution adopted by the present invention to solve the technical problem is: a method for tracking a target of a drone cluster based on an observer and a hierarchical network, comprising the following steps: 1) Divide the drones in a drone swarm into three roles: leader, follower, and observer; The leader is responsible for guiding the movement direction and speed of the entire cluster, planning the path according to the preset tasks or dynamic goals in the environment, and passing instructions to followers and observers; The follower forms a convex hull around the leader based on the preset reference point and adjusts its position and speed according to the leader's instructions. The relative position between the follower and the leader changes, but the distance remains unchanged; Observers are responsible for monitoring changes in the environment, including the appearance of obstacles or other potential threats. When an observer detects changes in the environment, it sends data to the leader and followers. The observer position is dynamically adjusted as the leader and follower clusters move, and is placed in a position that provides the maximum coverage of the surrounding environment of the drone cluster so that changes can be discovered in a timely manner; N unmanned aerial vehicles (UAVs) form a multi-agent system. In the system, each UAV acts as an independent agent. The multi-agent system is responsible for the formation of an orderly formation of the entire UAV group through the leader-follower-observer model. 2) Introducing a two-layer joint UAV network model to establish flying ad hoc networks (FANETs) responsible for communication in UAV cluster target tracking; Use SDN-based wireless defined network-based heterogeneous architecture for reliable communication in flight ad hoc networks; The drone swarm is divided into a controller, an upper drone layer, and a lower drone layer; the upper drone layer is responsible for performing mobility estimation and prediction operations and making routing decisions in conjunction with the controller, while the lower drone layer is responsible for performing ground or air search tasks or providing services; The upper drone layer includes the cluster head drone CHU, and the lower drone layer includes the cluster member drones CMU; one cluster head drone manages multiple cluster member drones; Each drone is equipped with a global navigation satellite system for obtaining drone location data; Calculate the Euclidean distance between drones; The position coordinates of UAV c and UAV i at time t; Each UAV communicates with neighboring UAVs through its UAV address identification, and the controller estimates the transmission routing cost of each UAV request. The controller calculates the optimal routing path for the cluster head UAV based on the global network information; 3) After the drone swarm realizes formation control, it can track the target. Based on the ant colony algorithm, it can simulate the collective behavior of ants to effectively track single or multiple targets. It can use the swarm intelligence characteristics of drones, information interaction and collaborative control to improve the accuracy and efficiency of tracking.

[0005] According to the above scheme, in step 1), each drone acts as an independent intelligent entity, equipped with a high-performance processor, sensor and wireless communication module.

[0006] According to the above scheme, in step 1), each drone acts as an independent intelligent entity and obtains its own position by being equipped with a navigation and positioning system.

[0007] According to the above scheme, in step 2), the Euclidean distance between the UAV c and the adjacent UAV i in the formation is expressed as:

[0008] in( , , )and( , , ) represent the position coordinates of UAV c and UAV i at time t respectively.

[0009] According to the above scheme, in step 2), the controller calculates the optimal routing path for the cluster head UAV based on the global network information; the details are as follows: The extended Kalman filter (EKF) is used to accurately estimate and predict the mobility of UAVs. The cluster head UAV CHUs control their corresponding cluster member UAVs CMUs and calculate and output the global optimal path through the DPSO routing decision algorithm to improve communication efficiency.

[0010] According to the above scheme, in step 2), each drone communicates with neighboring drones through its drone address identification, as follows: Through the collaborative work of the central controller and cluster head drones, routing decisions are made and communications are carried out; In the MAC architecture, the entire drone group shares spectrum resources through orthogonal frequency division multiple access (OFDMA). The cluster head drones (CHU) use IEEE 802.11p with a dedicated channel for H2H communication. The cluster head drone (CHU) and cluster member drones (CMU) use LTE for H2M communication. The cluster member drones (CMU) use IEEE802.11p that follows the EDCF standard for M2M communication to achieve high-quality communication connections.

[0011] According to the above scheme, in step 3), tracking of single or multiple targets is achieved by simulating the collective behavior of ants based on the ant colony algorithm.

[0012] According to the above scheme, in step 3), tracking of a single or multiple targets is achieved; 3.1) For single target tracking, the drone’s behavior strategy is as follows: When all drones fail to sense the target, they act in a cluster-controlled manner, moving toward a virtual target point according to predetermined rules or pre-set paths.

[0013] If some drones sense the target, they will start to adjust their behavior, maintain a certain distance from other drones, and move towards the sensed target position. Other drones that do not sense the target will adjust their behavior by observing the motion status of surrounding drones, and finally converge to achieve the purpose of consistent tracking.

[0014] 3.2) For multi-target tracking, the drone’s behavior strategy is as follows: When a drone senses a single target, the drone swarm quickly gathers and matches the target’s moving speed to adjust speed and track the target; When a drone senses multiple targets, the drone group quickly separates and forms different sub-groups, each of which tracks different targets. Other drones that have not sensed any targets will decide to voluntarily follow the sub-group with target information through information interaction.

[0015] The beneficial effects produced by the present invention are: 1. The present invention introduces the concept of observers to achieve effective formation control of the swarm. The leader, as the navigation core, leads the overall direction, and the followers follow the leader to perform tasks. At the same time, the observer nodes are reasonably arranged so that the drone group can have a more comprehensive understanding of the surrounding environment and make corresponding adjustments in time, so that the drone group can maintain a tighter and more orderly formation when performing tasks, improving the overall mobility and flexibility; 2. The present invention optimizes the communication loss problem in the drone cluster control network. By designing a two-layer joint drone network technology, it avoids problems such as inefficient path discovery, excessive hops, and lengthy paths. It makes the best routing decisions based on global information, improves communication efficiency, and achieves high-quality communication connections. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which: Figure 1 is a method flow chart of an embodiment of the present invention; Figure 2 A schematic diagram of a process of forming a convex hull between a leader and a follower according to an embodiment of the present invention; Figure 3 A flowchart of the steps of single and multiple target tracking according to an embodiment of the present invention; Figure 4 A schematic diagram of single target trajectory tracking according to an embodiment of the present invention; Figure 5 Schematic diagram of multi-target tracking and capture according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0018] like Figure 1 As shown, a method for tracking a target of a drone cluster based on an observer and a hierarchical network includes the following steps: 1) Divide the drones in a drone swarm into three roles: leader, follower, and observer; The leader is responsible for guiding the movement direction and speed of the entire cluster, planning the path according to the preset tasks or dynamic goals in the environment, and passing instructions to followers and observers; The follower forms a convex hull around the leader based on the preset reference point and adjusts its position and speed according to the leader's instructions. The relative position between the follower and the leader changes, but the distance remains unchanged; In this embodiment, Figure 2 , the follower takes the leader as the center point and changes into the desired geometric shape over time. The follower forms a convex hull around the leader and adjusts its position and speed according to the leader's instructions. Observers are responsible for monitoring changes in the environment, including the appearance of obstacles or other potential threats. When an observer detects changes in the environment, it sends data to the leader and followers. N unmanned aerial vehicles (UAVs) form a multi-agent system. In the system, each UAV acts as an independent agent, equipped with high-performance processors, sensors, and wireless communication modules to support complex network protocols and algorithm operations. It is equipped with the Beidou Navigation System (BNS) to obtain noisy positions. The multi-agent system introduces a leader-follower-observer model, which is responsible for the formation of an orderly formation of the entire UAV group; it implements the underlying control to ensure that the UAV cluster achieves the desired formation tracking when tracking the target; The concept of observers is introduced to achieve effective formation control of the swarm. The leader, as the navigation core, leads the overall direction, and the followers follow the leader to perform tasks. At the same time, the observer nodes are arranged reasonably, so that the drone group can have a more comprehensive understanding of the surrounding environment and make corresponding adjustments in time, so that the drone group can maintain a tighter and more orderly formation when performing tasks, improving the overall maneuverability and flexibility.

[0019] 2) Introducing a two-layer joint UAV network model to establish flying ad hoc networks (FANETs) responsible for communication in UAV cluster target tracking; Use SDN-based wireless defined network-based heterogeneous architecture for reliable communication in flight ad hoc networks; Divide the drone swarm into a controller, an upper drone layer, and a lower drone layer; The upper UAV layer includes the cluster head UAV CHU, and the lower UAV layer includes the cluster member UAV CMU; Each drone is equipped with a global navigation satellite system for obtaining drone location data; Calculate the Euclidean distance between drones; The Euclidean distance between UAV c and its adjacent UAV i in the formation is expressed as:

[0020] in( , , )and( , , ) represent the position coordinates of UAV c and UAV i at time t respectively.

[0021] Each UAV communicates with neighboring UAVs through its UAV address identification, and the controller estimates the transmission routing cost of each UAV request. The controller calculates the optimal routing path for the cluster head UAV based on the global network information; Through the collaborative work of the central controller and cluster head drones, routing decisions are made and communication is carried out. In the MAC architecture, the entire drone group shares spectrum resources through orthogonal frequency division multiple access (OFDMA). The cluster head drones (CHUs) use IEEE 802.11p with dedicated channels for H2H communication. The cluster head drones (CHUs) and cluster member drones (CMUs) use LTE for H2M communication. The cluster member drones (CMUs) use IEEE 802.11p that follows the EDCF standard for M2M communication to achieve high-quality communication connections. The extended Kalman filter (EKF) is used to accurately estimate and predict the mobility of drones. CHUs control their corresponding CMUs to select appropriate routing paths. The DPSO routing decision algorithm is used to calculate and output the global optimal path to improve communication efficiency.

[0022] 3) After the drone swarm realizes formation control, it can track the target and effectively track single or multiple targets. The swarm intelligence characteristics of drones are utilized to improve the accuracy and efficiency of tracking through information interaction and collaborative control.

[0023] In step 3), the tracking of single or multiple targets is achieved by simulating the collective behavior of ants based on the ant colony algorithm.

[0024] like Figure 3 As shown, it is divided into two cases: single target tracking and multi-target tracking: 3.1) For single target tracking, the drone’s behavior strategy is as follows: like Figure 4 As shown, when all drones fail to sense the target, the drones act in a cluster control manner and move toward a virtual target point according to predetermined rules or pre-set paths.

[0025] If some drones sense the target, they will start to adjust their behavior, maintain a certain distance from other drones, and move towards the sensed target position. Other drones that do not sense the target will adjust their behavior by observing the motion status of surrounding drones, and finally converge to achieve the purpose of consistent tracking.

[0026] 3.2) For multi-target tracking, the drone’s behavior strategy is as follows: like Figure 5 As shown, when a drone senses a single target, the drone swarm quickly gathers and matches the target's moving speed to adjust the speed to track the target; When a drone senses multiple targets, the drone group quickly separates and forms different sub-groups, each of which tracks different targets. Other drones that have not sensed any targets will decide to voluntarily follow the sub-group with target information through information interaction. It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A method for tracking target of drone cluster based on observer and hierarchical network, characterized in that: The following steps are involved: 1) Divide the drones in a drone swarm into three roles: leader, follower, and observer; The leader is responsible for guiding the movement direction and speed of the entire cluster, planning the path according to the preset tasks or dynamic goals in the environment, and passing instructions to followers and observers; The follower forms a convex hull around the leader based on the preset reference point and adjusts its position and speed according to the leader's instructions. The relative position between the follower and the leader changes, but the distance remains unchanged; Observers are responsible for monitoring changes in the environment, including the appearance of obstacles or other potential threats. When an observer detects changes in the environment, it sends data to the leader and followers. The observer position is dynamically adjusted as the leader and follower clusters move, and is placed in a position that provides the maximum coverage of the surrounding environment of the drone cluster so that changes can be discovered in a timely manner; N drones form a multi-agent system. In the system, each drone acts as an independent agent. The multi-agent system is responsible for the formation of an orderly formation of the entire drone group through the leader-follower-observer model. 2) Introducing a two-layer joint UAV network model to establish flying ad hoc networks (FANETs) responsible for communication in UAV cluster target tracking; Use SDN-based wireless defined network-based heterogeneous architecture for reliable communication in flight ad hoc networks; The drone swarm is divided into a controller, an upper drone layer, and a lower drone layer; the upper drone layer is responsible for performing mobility estimation and prediction operations and making routing decisions in conjunction with the controller, while the lower drone layer is responsible for performing ground or air search tasks or providing services; The upper drone layer includes the cluster head drone CHU, and the lower drone layer includes the cluster member drones CMU; one cluster head drone manages multiple cluster member drones; Calculate the Euclidean distance between drones; Each UAV communicates with neighboring UAVs through its UAV address identification, and the controller estimates the transmission routing cost of each UAV request. The controller calculates the optimal routing path for the cluster head UAV based on the global network information; 3) After the drone swarm achieves formation control, it conducts target tracking to achieve effective tracking of single or multiple targets.

2. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In step 1), each drone acts as an independent intelligent entity, equipped with a high-performance processor, sensor and wireless communication module.

3. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In step 1), each drone acts as an independent intelligent entity and obtains its own position by being equipped with a navigation and positioning system.

4. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In step 2), the Euclidean distance between the UAV c and the adjacent UAV i in the formation is expressed as: in( , , )and( , , ) represent the position coordinates of UAV c and UAV i at time t respectively.

5. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In step 2), the controller calculates the optimal routing path for the cluster head drone based on the global network information; specifically, as follows: The extended Kalman filter is used to accurately estimate and predict the mobility of UAVs. The cluster head UAVs (CHUs) control their corresponding cluster member UAVs (CMUs) and calculate and output the global optimal path through the DPSO routing decision algorithm to improve communication efficiency.

6. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In step 2), each drone communicates with neighboring drones through its drone address identification, as follows: Through the collaborative work of the central controller and cluster head drones, routing decisions are made and communications are carried out; In the MAC architecture, the entire drone group shares spectrum resources through orthogonal frequency division multiple access. The cluster head drones use IEEE 802.11p with a dedicated channel for H2H communication, the cluster head drone and cluster member drones use LTE for H2M communication, and the cluster member drones use IEEE 802.11p that follows the EDCF standard for M2M communication to achieve high-quality communication connections.

7. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In the step 3), tracking of a single or multiple targets is achieved by simulating the collective behavior of ants based on the ant colony algorithm.

8. The method for tracking target of drone cluster based on observer and hierarchical network according to claim 1 is characterized in that: In the step 3), tracking of a single or multiple targets is achieved; 3.1) For single target tracking, the drone’s behavior strategy is as follows: When all drones fail to sense the target, they act in a cluster-controlled manner, moving toward a virtual target point according to predetermined rules or pre-set paths. If some drones sense the target, they will start to adjust their behavior, keep a certain distance from other drones, and move towards the sensed target. Other drones that do not sense the target will adjust their behavior by observing the motion status of surrounding drones, and finally converge to achieve the purpose of consistent tracking. 3.2) For multi-target tracking, the drone’s behavior strategy is as follows: When a drone senses a single target, the drone swarm quickly gathers and matches the target’s moving speed to adjust speed and track the target; When a drone senses multiple targets, the drone group quickly separates and forms different sub-groups, each of which tracks different targets. Other drones that have not sensed any targets will decide to voluntarily follow the sub-group with target information through information interaction.

Images