A far and near cooperative perception unmanned aerial vehicle conflict obstacle avoidance method, device, equipment and medium

Abstract

The present application relates to the technical field of unmanned aerial vehicle obstacle avoidance, and discloses a near-far cooperative sensing unmanned aerial vehicle conflict obstacle avoidance method, device, equipment and medium, which is applied to an unmanned aerial vehicle conflict obstacle avoidance system, abnormal unmanned aerial vehicles are identified by first collecting and analyzing operation area information to predict their trajectory intrusion risks, trajectory data of ground and airborne radars are fused, and the vertical shortest distance of double-machine targets is accurately calculated, and obstacle avoidance processing is carried out on the basis thereof. The method realizes near-far cooperation of ground and airborne sensing, greatly improves the accuracy of trajectory data, makes distance calculation more in line with actual airspace situation, can accurately determine the threat level of abnormal unmanned aerial vehicles, provides reliable data support for operation unmanned aerial vehicle obstacle avoidance, effectively shortens the obstacle avoidance response time, improves the timeliness and effectiveness of unmanned aerial vehicle conflict obstacle avoidance in complex airspace, and guarantees the flight operation safety of operation unmanned aerial vehicles.

Description

Technical Field

[0001] This invention relates to the field of drone obstacle avoidance technology, specifically to a method, apparatus, equipment, and medium for drone conflict obstacle avoidance based on near and far collaborative perception. Background Technology

[0002] The development of drones is rapid, and they are now widely used in fields such as power line inspection, logistics delivery, and emergency rescue. Operational drones have become key equipment for improving production efficiency and reducing labor costs. However, the low-altitude airspace environment is complex, and the surge in data on abnormal drones, such as those illegally intruding or flying out of control, poses a significant threat to normal drone operations. In 2025 alone, there were multiple collisions between operational drones and abnormal drones. Therefore, drone collision avoidance has become one of the most pressing problems to be solved in the drone industry.

[0003] However, current perception technologies for collision avoidance in operational drones fall into two categories: single-sensor and multi-sensor fusion. Compared to single-sensor perception, multi-sensor fusion can compensate for the shortcomings of a single mode, improve the accuracy of detecting abnormal drones, and enhance adaptability to complex environments. However, current sensor fusion solutions, with their fixed parameter configuration of "radar / depth camera + ordinary camera," do not dynamically adjust the perception strategy for abnormal drones. This is especially problematic when abnormal drones rapidly and closely approach the scene, making it difficult to respond within a certain distance to resolve collisions and avoid obstacles. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for UAV conflict avoidance based on near-far collaborative perception, in order to solve the problem that the existing technology does not dynamically adjust the perception strategy for abnormal UAVs, especially when abnormal UAVs are rapidly and closely approaching the scene, making it difficult to complete the response within a certain distance to solve the conflict avoidance problem.

[0005] In a first aspect, the present invention provides a method for UAV conflict avoidance based on near-far cooperative sensing, applied to a UAV conflict avoidance system, the system comprising an operational UAV and a ground positioning platform; the method comprising: Collect information on the work area where the drone is performing its mission, and analyze whether there are any abnormal drones in the work area information; Based on the analysis results, predict whether the flight trajectory of the abnormal drone will enter the warning distance range of the operational drone; Based on the prediction results, the flight trajectory data of the abnormal UAVs detected by the airborne radar on the ground positioning platform and the operation UAV are extracted and fused. The shortest vertical distance between the abnormal drone and the operational drone is calculated by using the fused flight trajectory data of the abnormal drone and the operational drone. Based on the shortest vertical distance to the target, the operating drone is controlled to perform collision avoidance.

[0006] This invention is applied to UAV conflict and obstacle avoidance systems, establishing a complete link from "ground-based coarse positioning to airborne multimodal fine positioning to dynamic adjustment of early warning distance to conflict and obstacle avoidance." It first collects and analyzes information about the operational area to identify abnormal UAVs and predict their trajectory intrusion risk. Then, it integrates trajectory data from ground and airborne radar to accurately calculate the shortest vertical distance between two targets, using this as the basis for obstacle avoidance. This method achieves near-far coordination between ground and airborne perception, significantly improving the accuracy of trajectory data, making distance calculations more consistent with the actual airspace situation, accurately determining the threat level of abnormal UAVs, providing reliable data support for obstacle avoidance of operational UAVs, effectively shortening obstacle avoidance response time, improving the timeliness and effectiveness of UAV conflict and obstacle avoidance in complex airspace, and ensuring the flight safety of operational UAVs.

[0007] In one optional implementation, the step of collecting information about the work area where the drone is performing its task and monitoring whether there are any abnormal drones in the work area information includes: The ground positioning platform collects information about the operational area where the drone is performing its mission. Analyze the information in the work area to determine if there are any abnormal drones; If not, proceed to the step of collecting information about the work area where the UAV is performing its mission through the ground positioning platform.

[0008] This invention utilizes a ground-based positioning platform to continuously collect information about the operational area and detect abnormal drones. When no abnormalities are detected, monitoring is performed cyclically, enabling normalized and uninterrupted awareness of the operational airspace. Leveraging the global detection advantages of the ground-based positioning platform, it ensures full coverage and continuity of airspace monitoring, and can detect abnormal drone intrusion signals in real time.

[0009] In one optional implementation, the step of predicting whether the flight trajectory of the abnormal UAV enters the warning distance range of the operational UAV based on the analysis results includes: When the analysis results indicate the presence of an abnormal drone in the work area information, the flight parameters of the abnormal drone and the flight parameters of the work drone are extracted. Using the flight parameters of the abnormal UAV and the flight parameters of the operational UAV, calculate the initial vertical shortest distance between the abnormal UAV and the operational UAV; Determine whether the initial vertical shortest distance is less than or equal to a preset warning distance threshold and greater than a preset alert distance threshold; If so, it is determined that the flight trajectory of the abnormal drone has entered the warning distance range of the operational drone.

[0010] This invention, upon identifying an abnormal drone, extracts the flight parameters of both drones and calculates the initial shortest vertical distance. By comparing this distance with a warning distance threshold, it determines the risk of trajectory intrusion. Relying on precise parameter extraction and distance calculation, it achieves rapid quantitative assessment of the threat posed by abnormal drones. This assessment method is logically direct and computationally efficient, accurately identifying warning-level threats and promptly triggering subsequent processes, significantly reducing threat assessment time and improving the early warning response efficiency of drone conflict and obstacle avoidance.

[0011] In one optional implementation, the step of extracting and fusing the flight trajectory data of the abnormal UAVs detected by the ground positioning platform and the airborne radar on the operational UAV based on the prediction results includes: When the flight path of the abnormal drone enters the warning range of the operational drone, the onboard radar and high-speed camera of the operational drone are activated. The airborne radar detects the flight parameters of the abnormal drone, and the high-speed camera captures image data of the drone. By combining the flight parameters detected by the airborne radar and the image data, close-range flight trajectory data of the abnormal drone is generated; Extract the global flight trajectory data of the abnormal UAV detected by the ground positioning platform; The close-range flight trajectory data and the global flight trajectory data are fused together.

[0012] This invention activates airborne radar and high-speed cameras immediately upon an abnormal drone entering the warning range, fusing the first flight trajectory from airborne multi-source sensing with the second flight trajectory from the ground positioning platform to achieve complementary fusion of near and far sensing data. By leveraging the global detection advantages of the ground platform and combining the precise near-range sensing capabilities of the airborne equipment, it significantly improves the accuracy and completeness of abnormal drone trajectory data.

[0013] In one optional implementation, the fusion processing of the near-field flight trajectory data and the global flight trajectory data includes: The near-field flight trajectory data and the global flight trajectory data are preprocessed; The preprocessed near-field flight trajectory data and global flight trajectory data are timestamped and aligned. Using the coordinate positions corresponding to the timestamp-aligned close-range flight trajectory data and the global flight trajectory data as the initial state of the Kalman filter, the timestamp-aligned close-range flight trajectory data and the global flight trajectory data are fused using the Kalman filter fusion iterative method.

[0014] This invention preprocesses dual-track data and then performs precise timestamp alignment. Using the aligned coordinates as the initial state for Kalman filtering, trajectory fusion is achieved through iterative filtering and fusion. Preprocessing effectively reduces data noise interference, timestamp alignment ensures data temporal consistency, and Kalman filtering enables precise fusion of near and far trajectory data, significantly improving the accuracy and real-time performance of abnormal UAV trajectory data.

[0015] In one optional implementation, controlling the operational UAV to perform collision avoidance based on the shortest vertical distance to the target includes: Determine whether the shortest vertical distance to the target is less than or equal to a preset warning distance threshold; If the distance is less than or equal to the preset warning distance threshold, then determine whether the shortest vertical distance to the target is less than or equal to the preset countermeasure distance threshold. If the distance exceeds the preset countermeasures threshold, a warning message is issued to the abnormal drone. If the distance is less than or equal to the preset countermeasure distance threshold, then the corresponding countermeasure operation is performed according to the type of the abnormal drone, and the operation drone is controlled to perform conflict avoidance processing.

[0016] This invention achieves differentiated alarm, countermeasure, and obstacle avoidance handling by classifying the relationship between the target's shortest vertical distance and alarm / countermeasure thresholds. It first triggers an alarm to warn of potential risks, then, for countermeasure-level threats, executes specific countermeasures based on the drone type while simultaneously avoiding obstacles. This clear hierarchy and precise handling not only achieves a tiered response to threats but also makes countermeasures more targeted, effectively improving the rationality and effectiveness of obstacle avoidance and countermeasures, and minimizing the collision risk of operational drones.

[0017] In an optional implementation, the method further includes: When the abnormal drone enters the warning distance range and the type of the abnormal drone is a micro drone, countermeasures are performed on the abnormal drone by ground-based countermeasure equipment. When the abnormal drone enters the warning distance range and the type of the abnormal drone is a light drone, the ground countermeasure equipment performs a countermeasure operation on the abnormal drone. When the abnormal drone is a small drone and the collision time between the flight trajectory data of the abnormal drone and the flight trajectory data of the operational drone is less than a preset time threshold, the operational drone is controlled to perform collision avoidance processing.

[0018] This invention implements tiered ground countermeasures and airborne obstacle avoidance based on the target's shortest vertical distance threshold and the type of abnormal drone. Micro and light drones trigger ground countermeasures at warning and alert distances, respectively, while small drones initiate obstacle avoidance based on a collision time threshold. This strategy achieves precise adaptation between countermeasures and obstacle avoidance, triggering ground / airborne response measures as needed, thus improving the targeting and efficiency of conflict response.

[0019] Secondly, the present invention provides a remote-to-remote collaborative sensing UAV conflict avoidance device, applied to a UAV conflict avoidance system, the system comprising an operational UAV and a ground positioning platform; the device comprising: The data acquisition module is used to collect information about the work area where the drone is performing its mission, and to analyze whether there are any abnormal drones in the work area information. The prediction module is used to predict, based on the analysis results, whether the flight trajectory of the abnormal drone will enter the warning distance range of the operational drone; The fusion module is used to extract and fuse the flight trajectory data of the abnormal UAVs detected by the ground positioning platform and the airborne radar on the operation UAV, respectively, based on the prediction results. The calculation module is used to calculate the shortest vertical distance between the abnormal drone and the operational drone by using the fused flight trajectory data of the abnormal drone and the flight trajectory data of the operational drone. The control module is used to control the operating drone to perform collision avoidance based on the shortest vertical distance to the target.

[0020] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the near-far cooperative perception UAV conflict avoidance method described in the first aspect or any corresponding embodiment.

[0021] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the near-far cooperative perception UAV conflict avoidance method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a schematic flowchart of the first type of UAV conflict avoidance method based on near-far cooperative perception according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the composition of an unmanned aerial vehicle (UAV) collision avoidance system according to an embodiment of the present invention; Figure 3 This is a basic structural diagram of an operational unmanned aerial vehicle according to an embodiment of the present invention; Figure 4 This is a flowchart illustrating the visual dynamic scheduling strategy according to an embodiment of the present invention; Figure 5 This is a second flowchart illustrating the UAV conflict avoidance method based on near-far cooperative perception according to an embodiment of the present invention. Figure 6 This is a structural block diagram of a UAV conflict avoidance device with near-far cooperative sensing according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention.

[0024] The reference numerals in the attached figures are explained as follows: 1. Ground station; 2. Operational drone; 3. Ground radar; 4. Abnormal drone; 21. Countermeasure device; 22. LiDAR; 23. High-speed camera; 24. Operational equipment. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0027] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0028] This invention provides a method for UAV conflict avoidance based on near-far collaborative perception, establishing a complete link of "ground positioning platform coarse positioning - airborne multimodal fine positioning - dynamic adjustment of early warning distance - conflict avoidance". First, it collects and analyzes information about the operational area to identify abnormal UAV 4 and predict its trajectory intrusion risk. Then, it fuses trajectory data from ground and airborne radar to accurately calculate the shortest vertical distance between the two targets, using this as the basis for obstacle avoidance. This method achieves near-far collaboration between ground and airborne perception, significantly improving the accuracy of trajectory data, making distance calculations more consistent with the actual airspace situation, and accurately determining the threat level of abnormal UAV 4. This provides reliable data support for obstacle avoidance by the operational UAV 2, effectively shortening the obstacle avoidance response time, improving the timeliness and effectiveness of UAV conflict avoidance in complex airspace, and ensuring the flight safety of the operational UAV 2.

[0029] According to an embodiment of the present invention, a method for UAV conflict avoidance based on near-far cooperative perception is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This embodiment provides a UAV conflict avoidance method based on near and far collaborative perception, which is applied to a UAV conflict avoidance system. The system includes an operational UAV 2 and a ground positioning platform. Figure 1 This is a flowchart of a UAV conflict avoidance method based on near-far cooperative perception according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Collect information on the work area where the operation drone 2 is performing the task, and analyze whether there is an abnormal drone 4 in the work area information.

[0031] It should be noted that the UAV conflict avoidance system includes the operational UAV 2 and a ground positioning platform; the ground positioning platform includes a ground radar 3 and a ground station 1. Specifically, ground station 1 is used to monitor the operational status and flight parameters of the operational UAV 2; ground radar 3 is used to monitor the operational area of ​​the operational UAV 2 using phased array radar. The top of the operational UAV 2 is equipped with a lidar 22 (i.e., airborne radar) and a high-speed camera 23, while the bottom is equipped with operational equipment 24 and a countermeasure device 21. The operational equipment 24 can be surveying instruments, communication measuring instruments, etc.

[0032] Operational area information refers to the collective term for various data that reflect the overall situation of the airspace, such as the spatial coordinates, flight speed, and flight direction of all flying targets within the designated airspace where the operational UAV 2 performs its mission.

[0033] Abnormal drone 4 refers to non-operational drone 2 that is not included in the operational airspace planning, illegally intrudes, or flies out of control.

[0034] In this embodiment of the invention, the ground radar 3 of the ground positioning platform continuously detects and acquires information about the operation area. This information includes airspace situation data such as the spatial coordinates, flight speed, and flight direction of all flying targets in the operation airspace. The ground positioning platform transmits the detected airspace situation data to the ground station 1. The ground station 1 analyzes the airspace situation data through a target recognition algorithm, filters out flying targets that match the characteristics of UAVs, and further determines whether the flying target is an abnormal UAV 4 that is not part of the planned operation.

[0035] It is worth mentioning that the operational drone 2, such as Figure 3 As shown, the core of the entire system is the drone equipped with a measurement unit or actuator, such as a UAV carrying a spectrum analyzer for low-altitude communication detection. Typically, the operational UAV 2 operates within a designated area along a designated path. To enable the operational UAV 2 to detect abnormal UAV 4, it needs to be equipped with a lidar 22 and a high-speed camera 23 for multimodal collision detection, while also supporting lightweight countermeasure equipment for emergency countermeasures.

[0036] Ground station 1 monitors the operational status and flight parameters of the operational drone 2. Simultaneously, it connects to ground radar 3 to acquire current airspace information, especially the flight trajectory, coordinates, and flight parameters of the abnormal drone 4. Ground radar 3 establishes a real-time warning distance for the operational drone 2 based on its data and synchronizes this information to the drone. If the abnormal drone 4 exceeds the warning range, ground station 1 triggers a command to instruct the operational drone 2 to activate lidar 22 and high-speed camera 23 to capture the flight status of the abnormal drone 4. Ground station 1 fuses the flight trajectory monitored by ground radar 3 with the flight trajectory captured by airborne sensors to perform precise situational awareness of the illegal drone and determine whether a collision with the operational drone 2 is imminent.

[0037] Ground-based radar 3, employing a phased array radar, is used to monitor the UAV's operational area, constructing a "long-range early warning + near-range avoidance" defense strategy. Ground-based radar 3 is generally large and heavy, typically deployed on a mobile platform. Compared to the airborne lidar 22, ground-based radar 3 has a longer detection range and wider coverage, providing global airspace situational awareness for the operational UAV 2. After activation, ground-based radar 3 monitors the operational airspace. When an abnormal UAV 4 intrudes into the operational airspace, ground station 1 analyzes its flight trajectory. When the predicted flight trajectory of abnormal UAV 4 enters the early warning range of operational UAV 2, an early warning command is sent to operational UAV 2, activating its onboard lidar 22 + high-speed camera 23.

[0038] Step S102: Based on the analysis results, predict whether the flight trajectory of the abnormal drone 4 will enter the warning distance range of the operational drone 2.

[0039] It should be noted that the flight trajectory refers to the path formed by the spatial coordinates of the UAV at different time points during flight, according to the flight sequence.

[0040] The warning distance range refers to the preset safe airspace distance range threshold for the operational drone 2, which is the critical distance that triggers the activation of the airborne radar and high-speed camera 23 after the abnormal drone 4 intrudes.

[0041] In this embodiment of the invention, when the ground station 1 determines that there is an abnormal drone 4 in the work area, it immediately extracts the real-time flight parameters of the abnormal drone 4 and the current flight parameters of the work drone 2, and uses the trajectory prediction algorithm to predict the subsequent flight trajectory of the abnormal drone 4. At the same time, combined with the work route of the work drone 2, it calculates the trend of the change of the target vertical shortest distance between the predicted trajectory and the flight trajectory of the work drone 2, and predicts whether the flight trajectory of the abnormal drone 4 will intrude into the preset warning distance range for the work drone 2.

[0042] Step S103: Extract the flight trajectory data of the abnormal UAV 4 detected by the airborne radar on the ground positioning platform and the operation UAV 2 respectively based on the prediction results, and perform fusion processing.

[0043] It should be noted that the operation drone 2 is equipped with airborne radar and high-speed camera 23. The airborne radar refers to the radar detection equipment mounted on the operation drone 2, which has the characteristics of short-range and high-precision detection and can capture the flight trajectory data of abnormal drones 4 around the operation drone 2 in real time.

[0044] Flight trajectory data refers to the core data set that reflects the flight trajectory of a drone, including key information such as the spatial coordinates, flight speed, and trajectory direction corresponding to four different timestamps of the abnormal drone.

[0045] Fusion processing refers to the process of integrating and optimizing multi-source data of the same target acquired by different detection devices through a Kalman filter fusion iterative algorithm.

[0046] In this embodiment of the invention, when the ground station 1 predicts that the flight trajectory of the abnormal UAV 4 will enter the warning range of the operation UAV 2, it immediately triggers a data extraction command, and simultaneously retrieves the global flight trajectory data of the abnormal UAV 4 detected by the ground positioning platform, as well as the close-range flight trajectory data of the abnormal UAV 4 detected by the airborne radar of the operation UAV 2. The two types of trajectory data are transmitted to the ground station 1, and multi-source trajectory fusion processing is performed on them by the Kalman filter fusion iterative method to form accurate and unified fused flight trajectory data of the abnormal UAV 4.

[0047] Step S104: Using the fused flight trajectory data of the abnormal UAV 4 and the flight trajectory data of the operational UAV 2, calculate the shortest vertical distance between the abnormal UAV 4 and the operational UAV 2.

[0048] It should be noted that the shortest vertical distance between the target refers to the shortest straight-line distance in the vertical height dimension of the three-dimensional airspace between the flight trajectories of the abnormal UAV 4 and the operational UAV 2.

[0049] In this embodiment of the invention, after the ground station 1 obtains the accurate and unified fused flight trajectory data of the abnormal UAV 4, it simultaneously retrieves the real-time flight trajectory data of the operation UAV 2, aligns the two sets of trajectory data in time sequence based on the same timestamp, extracts the spatial coordinates of the two UAVs at each time node after alignment, and solves the shortest straight-line distance between the spatial coordinates of the two UAVs in the vertical dimension through the spatial distance calculation model, thereby obtaining the target shortest vertical distance between the abnormal UAV 4 and the operation UAV 2 at that time node.

[0050] Step S105: Based on the shortest vertical distance to the target, control the operation drone 2 to perform conflict avoidance.

[0051] It should be noted that conflict avoidance and obstacle handling refers to a series of operations performed by the operating UAV 2 after sensing the collision threat of the abnormal UAV 4 in the airspace, such as adjusting its flight status and activating the countermeasure device 21, to avoid collision with the abnormal UAV 4 and eliminate airspace conflict.

[0052] In this embodiment of the invention, the ground station 1 compares the calculated shortest vertical distance to the target with preset warning, alarm, and countermeasure distance thresholds to determine the threat level of the current abnormal drone 4 to the operation drone 2. Based on the threat level and the identified type of abnormal drone 4, the ground station 1 triggers the corresponding graded disposal command and sends it to the operation drone 2. After receiving the command, the operation drone 2 synchronously adjusts its own flight parameters and starts the relevant equipment to perform the appropriate obstacle avoidance flight or cooperative countermeasure operation to complete the conflict and obstacle avoidance processing to avoid the risk of collision.

[0053] It is worth mentioning that, such as Figure 2 and Figure 3 As shown, before the UAV performs its mission, the operational UAV 2 deploys an airborne platform equipped with a lidar 22 and a high-speed camera 23. The ground positioning platform is a mobile platform, which deploys a ground radar 3 and a ground station 1. During mission execution, the ground radar 3 is activated and monitors the current airspace, while the lidar 22 and high-speed camera 23 on the airborne platform are in standby mode. When the ground radar 3 detects an abnormal UAV 4, it analyzes the spatial coordinates and flight trajectory of the abnormal UAV 4. Once the abnormal UAV 4 enters the warning range of the operational UAV 2, the lidar 22 and high-speed camera 23 on the airborne platform are activated and capture the abnormal UAV 4. The flight trajectory obtained by the ground radar 3 and the flight trajectory perceived by the airborne multi-sensor are fused to analyze whether the abnormal UAV 4 is aggressive.

[0054] When an abnormal drone 4 rapidly approaches the operational drone 2, a visual dynamic scheduling strategy needs to be established based on the sensing capabilities of the operational drone 2's LiDAR 22 and high-speed camera 23 to address the challenge of rapid, close-range collision detection. Simultaneously, when the abnormal drone 4 enters the warning range, a response mechanism needs to be established to reduce physical damage from direct collisions. The visual dynamic scheduling strategy is a visual response of the operational drone 2 to potential collisions, ensuring the safety of the drone's fuselage and onboard equipment. The visual dynamic scheduling strategy includes a normal phase, a warning phase, a alert phase, and a countermeasure phase. The data and model strategies of its LiDAR 22 and high-speed camera 23 are as follows: Figure 4 As shown.

[0055] This embodiment provides a UAV conflict avoidance method based on near and far collaborative perception, which is applied to a UAV conflict avoidance system. The system includes an operational UAV 2 and a ground positioning platform. Figure 5 This is a flowchart of a UAV conflict avoidance method based on near-far cooperative perception according to an embodiment of the present invention, such as... Figure 5 As shown, the process includes the following steps: Step S201: Collect information on the work area where the operation drone 2 is performing the task, and analyze whether there is an abnormal drone 4 in the work area information.

[0056] Specifically, step S201 includes: Step S2011: Collect information on the work area where the UAV 2 is performing the task through the ground positioning platform.

[0057] In this embodiment of the invention, the ground positioning platform is pre-deployed around the operational airspace and completes communication adaptation with the operational drone 2. The lidar 22 on it continuously performs full-domain scanning and detection of the designated airspace where the operational drone 2 performs its mission, and captures in real time the airspace situation data such as the spatial position, flight speed, and flight direction of all flying targets in the airspace. After integrating these data, complete operational area information is formed and transmitted to the ground station 1. The flight status and operational status of the operational drone 2 are also transmitted to the ground station 1 at the same time.

[0058] Step S2012: Analyze the information of the work area to see if there are any abnormal drones 4.

[0059] In this embodiment of the invention, ground station 1 obtains information about the current work area and monitors in real time whether there is any abnormal drone 4 intruding into the work area.

[0060] If not, proceed to step S2013, then skip to the step of collecting information on the work area where the UAV 2 is performing its mission via a ground positioning platform.

[0061] In this embodiment of the invention, when no abnormal drone 4 intrudes into the work area, the information on the work area where the work drone 2 is located when performing the task is collected in real time.

[0062] Step S202: Based on the analysis results, predict whether the flight trajectory of the abnormal drone 4 will enter the warning distance range of the operational drone 2.

[0063] In some optional implementations, step S202 above includes: Step S2021: When the analysis result shows that there is an abnormal UAV 4 in the work area information, extract the flight parameters of the abnormal UAV 4 and the flight parameters of the work UAV 2.

[0064] It should be noted that flight parameters refer to the core data set of the UAV's flight status, including key information such as the UAV's real-time spatial coordinates, flight speed, flight altitude, and flight direction.

[0065] In this embodiment of the invention, when the abnormal drone 4 intrudes into the work area, the ground radar 3 detects the flight trajectory, coordinates, and flight parameters of the abnormal drone 4, and establishes the warning distance of the work drone 2 in real time based on the work drone 2, and synchronizes the relevant information to the work drone 2.

[0066] When the abnormal drone 4 rapidly approaches the operational drone 2, a visual dynamic scheduling strategy needs to be established based on the perception of the operational drone 2's lidar 22 and high-speed camera 23. Specifically, the visual dynamic scheduling strategy includes a normal phase, an early warning phase, a warning phase, and a countermeasure phase. The normal phase refers to a situation where there is no abnormal drone 4 in the current operational airspace or the flight path of the abnormal drone 4 predicted by the ground radar 3 does not pose a threat. The safe distance range in the normal phase is greater than 100 meters, which is the shortest vertical distance between the predicted flight path of the abnormal drone 4 and the operational drone 2. At this safe distance, the airborne radar and high-speed camera 23 are in standby mode, consuming no onboard computing power or battery power.

[0067] Step S2022: Using the flight parameters of the abnormal UAV 4 and the flight parameters of the operational UAV 2, calculate the initial vertical shortest distance between the abnormal UAV 4 and the operational UAV 2.

[0068] It should be noted that the initial vertical shortest distance refers to the first version of the vertical shortest distance calculated based on the real-time flight parameters of the two aircraft.

[0069] In this embodiment of the invention, after the ground station 1 synchronously extracts and summarizes the real-time flight parameters of the two aircraft, it calibrates the parameter timing based on the unified timestamp of the system, extracts the altitude coordinate values ​​of the corresponding time from the flight parameters of the two aircraft, and calculates the initial vertical shortest distance between the abnormal UAV 4 and the operation UAV 2 at that time node by performing numerical difference calculation and taking the absolute value.

[0070] Step S2023: Determine whether the initial vertical shortest distance is less than or equal to the preset warning distance threshold and greater than the preset warning distance threshold.

[0071] It should be noted that the preset warning distance threshold refers to the vertical distance critical value preset for the safe operation of the operational drone 2, which is the warning stage of the visual dynamic scheduling strategy. The warning stage refers to the situation where the flight path of the abnormal drone 4 predicted by the ground radar 3 constitutes a potential threat. The distance range of the warning stage is less than 100 meters and greater than 30 meters, that is, the shortest vertical distance between the flight path of the abnormal drone 4 predicted by the ground radar 3 and the operational drone 2. At this warning distance, the airborne radar and high-speed camera 23 are activated, with the camera set to a frame rate of 60fps and a resolution of 720p.

[0072] The preset warning distance threshold refers to a vertical distance threshold set for the safe flight of the operational UAV 2, which is higher than the countermeasure threshold but lower than the warning threshold. This threshold represents the warning phase of the visual dynamic scheduling strategy. Specifically, the warning phase occurs when the ground radar 3 and airborne radar fusion sensing indicate that the abnormal UAV 4 has entered the warning distance range and its continued flight will pose a threat to the operational UAV 2. The distance range for the warning phase is less than 30 meters and greater than 5 meters, which is the shortest vertical distance between the flight trajectories of the abnormal UAV 4 predicted by the ground radar 3 and the airborne radar, and the operational UAV 2. During this warning phase, the parameters of the airborne radar and high-speed camera 23 are set, with the lidar 22 undergoing 2 / 3 (adjustable) downsampling, and the camera set to 120fps and 640p resolution.

[0073] In this embodiment of the invention, after calculating the initial vertical shortest distance between the abnormal UAV 4 and the operational UAV 2, the ground station 1 accurately compares the real-time quantified value with the pre-set warning distance threshold of the system, and determines whether the initial vertical shortest distance has reached the warning stage triggering condition through logical judgment, thereby judging whether the abnormal UAV 4 has posed a potential airspace threat to the operational UAV 2.

[0074] Step S2024: If yes, then determine that the flight trajectory of the abnormal drone 4 has entered the warning distance range of the working drone 2.

[0075] In this embodiment of the invention, after confirming that the initial vertical shortest distance meets the warning triggering conditions, the ground station 1 determines that the flight trajectory of the abnormal drone 4 has entered the warning distance range of the working drone 2, and immediately generates a valid warning identifier, marking the abnormal drone 4 as a threat target that has intruded into the warning area.

[0076] Step S203: Extract the flight trajectory data of the abnormal UAV 4 detected by the airborne radar on the ground positioning platform and the operation UAV 2 respectively based on the prediction results, and perform fusion processing.

[0077] In some optional implementations, step S3022 above includes: Step S2031: When the flight path of the abnormal drone 4 enters the warning range of the operation drone 2, the airborne radar and high-speed camera 23 of the operation drone 2 are activated.

[0078] In this embodiment of the invention, when the flight trajectory of the abnormal UAV 4 enters the warning distance of the operational UAV 2, it serves as a trigger signal for the system to enter subsequent obstacle avoidance processes such as multi-source trajectory fusion, activation of airborne radar and high-speed camera 23, ensuring that the UAV conflict obstacle avoidance system can respond in a timely manner and intervene quickly.

[0079] In step S2032, the flight parameters of the abnormal drone 4 are detected by airborne radar, and the image data of the drone are captured by high-speed camera 23.

[0080] In this embodiment of the invention, image data refers to continuous frame images captured by the high-speed camera 23, which include visual feature information such as the appearance, outline, and relative position of the abnormal drone 4.

[0081] The combination of LiDAR 22 and high-speed camera 23 is more effective than LiDAR 22 plus a regular camera in detecting high-speed, dynamic, and anomalous drones 4. High-speed camera 23 can capture high-speed moving targets at a high frame rate, clearly capturing the target's dynamic trajectory. By establishing a visual dynamic scheduling strategy, the recognition speed of the airborne computing platform meets the perception requirements of the operational drone 2. A target recognition framework of "long-range recognition + short-range obstacle avoidance" is established. YOLOv11-Tiny (or other lightweight models) is used for identifying anomalous drones 4 at medium and long ranges. Combined with the flight data of anomalous drones 4 acquired by LiDAR 22, the category, size, and trajectory of anomalous drones 4 are obtained. When entering close range, a monocular estimation model of anomalous drones 4 is established using parameters from high-speed camera 23, simplifying the calculations of the airborne computing platform and enabling high-speed acquisition of the spatial position of anomalous drones 4.

[0082] Step S2033: Combine the flight parameters and image data of the abnormal UAV 4 detected by the airborne radar to generate close-range flight trajectory data of the abnormal UAV 4.

[0083] It should be noted that the close-range flight trajectory data refers to the abnormal drone trajectory data generated based on the detection data of the onboard equipment (airborne radar, high-speed camera 23) of the operational drone 2.

[0084] In this embodiment of the invention, point cloud data obtained from the flight parameters of the abnormal drone 4 detected by the lidar 22 and image data from the high-speed camera 23 are used to improve the accuracy of target recognition. Specifically, PointNet / PointCNN is used to extract the geometric features of the target point cloud on the point cloud side, while YOLOv11-Tiny is used to extract visual texture features on the image side. The feature data from both modalities are weighted and fused before being input into a classifier for target recognition. Based on the bounding box and background obtained from the recognition results, the target point cloud data is further segmented. Enclosure analysis is performed on the segmented point cloud data to obtain its centroid coordinates. By combining continuous point cloud data and image frames, the close-range flight trajectory data of the abnormal drone 4 is obtained, thus yielding the target flight trajectory coordinates.

[0085] Step S2034: Extract the global flight trajectory data of the abnormal UAV 4 detected by the ground positioning platform.

[0086] It should be noted that global flight trajectory data refers to trajectory data detected by the ground positioning platform, reflecting the flight path and status of the abnormal UAV 4 over a large area and a long period of time in the entire operational airspace.

[0087] In this embodiment of the invention, after generating the close-range flight trajectory data of the abnormal UAV 4, the ground station 1 simultaneously filters and extracts the global flight trajectory data corresponding to the same abnormal UAV 4 from the historical and real-time detection data of the ground positioning platform. This data covers the complete flight path and global position information of the abnormal UAV 4 since it entered the operating airspace, and complements the close-range flight trajectory data.

[0088] Step S2035: The near-field flight trajectory data and the global flight trajectory data are fused together.

[0089] In this embodiment of the invention, after acquiring the close-range flight trajectory data and global flight trajectory data of the same abnormal UAV 4, the ground station 1 performs data fusion processing through the Kalman filter iterative algorithm.

[0090] Specifically, step S2035 includes: Step a1: Preprocess the close-range flight trajectory data and the global flight trajectory data.

[0091] It should be noted that, regarding the monitoring of the flight trajectory of the anomalous UAV 4 by ground station 1, ground radar 3 can provide the global trajectory (i.e., global flight trajectory data), while airborne radar provides the close-range precision point cloud trajectory (i.e., close-range flight trajectory data). Therefore, the global flight trajectory data of the anomalous UAV 4 observed by ground radar 3 and the close-range flight trajectory data of the anomalous UAV 4 observed by airborne radar are fused using Kalman filtering. Since the computational cost of Kalman filtering is relatively small (<10ms), it can meet the observation time requirements for collision avoidance.

[0092] In this embodiment of the invention, the raw data is preprocessed as follows: Kalman filtering is sensitive to noise, and linearized noise is filtered through preprocessing. (a) Ground radar 3: The moving average filter is used to smooth random noise and outputs linearized data of "position, velocity, and UTC timestamp"; (b) LiDAR 22: Voxel grid is used for downsampling, the target point cloud is segmented by Euclidean clustering of the point cloud and its centroid coordinates are calculated, and "centroid coordinates and local timestamp" are output.

[0093] Step a2 involves aligning the preprocessed near-field flight trajectory data and global flight trajectory data with timestamps.

[0094] It should be noted that timestamp alignment refers to the operation of performing time-series calibration on two types of trajectory data: preprocessed near-field flight trajectory data and global flight trajectory data.

[0095] In this embodiment of the invention, the Kalman filter algorithm requires that the observation data be fused at the same time, retaining radar / liquid radar 22 data within ±1ms, and using the UTC timestamp of the flight control of the UAV 2 as the unique reference of the system. The timestamp alignment method is as follows: 1) Timestamp correction for ground radar 3: Record the wireless transmission delay τ of ground radar 3 data. Radar reference time = radar local time + τ. 2) The local timestamp of LiDAR 22 uses linear interpolation alignment. If the data timestamp of LiDAR 22 is... t 1. t 2. Kalman fusion frame time is t Then, at time 22, the lidar 22 t The coordinates are:

[0096] In the formula, This indicates that the lidar 22 is at time 22. The coordinates; This indicates that the lidar 22 is at time 22. The coordinates; t Indicates the Kalman fusion frame time; t 1. t 2 represents the timestamp of the LiDAR 22 data.

[0097] Step a3: Using the coordinate positions corresponding to the timestamp-aligned close-range flight trajectory data and the global flight trajectory data as the initial state of Kalman filtering, the timestamp-aligned close-range flight trajectory data and the global flight trajectory data are fused using the Kalman filter fusion iteration method.

[0098] It should be noted that the initial state of the Kalman filter refers to the initial input data when the Kalman filter algorithm starts.

[0099] Kalman filter fusion iterative method refers to an optimization algorithm for multi-source data fusion, which follows the logic of initialization (first time / reset) → [prediction step → observation fusion step → gain calculation step → update step → result caching].

[0100] In this embodiment of the invention, the initial state estimation is performed as follows: the ground radar 3 provides mid-to-long-range early warning, the ground station 1 sends a command to the operation drone 2 to start the lidar 22, and the abnormal drone 4 target is obtained by combining multimodal reasoning to obtain the flight coordinates of the abnormal drone 4, and the Kalman filter is used to initialize the state vector.

[0101] 1) Initialization: Set the initial state to provide starting data for the first Kalman filter: , Input the initial radar / lidar 22 coordinates after synchronization; 2) Prediction step, basic formula:

[0102]

[0103] In the formula, , Output Indicates the prior state. Indicates the prior covariance. express( k -1) The prior covariance matrix at time t.

[0104] 3) Observation fusion, basic formula:

[0105]

[0106] In the formula, Indicates the coordinates after synchronization; Indicates distance weight; , Output Indicates the merged observations; This indicates observation noise.

[0107] 4) Gain calculation, basic formula:

[0108] In the formula, , Indicates the prior covariance. Represents observation noise, output It is the Kalman gain.

[0109] 5) State update, basic formula:

[0110]

[0111] In the formula, the output It is the optimal trajectory. It is the posterior covariance. , Indicates the prior covariance. It is Kalman gain. Indicates the prior state. denoted as fused observations, and I represents the identity matrix with the same dimension as the state vector.

[0112] 6) Result caching: Cache the best result for the current frame. and , The fused trajectory is used for conflict and obstacle avoidance decision-making of the operational drone.

[0113] 7) Data anomaly handling: If lidar 22 has no observations, update it only with observations from ground radar 3; if neither has any observations, continue prediction until observations are recovered.

[0114] Step S204: Using the fused flight trajectory data of the abnormal UAV 4 and the flight trajectory data of the operational UAV 2, calculate the shortest vertical distance between the abnormal UAV 4 and the operational UAV 2.

[0115] Please see details Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0116] Step S205: Based on the shortest vertical distance to the target, control the operation drone 2 to perform conflict avoidance.

[0117] In some optional implementations, step S205 above includes: Step S2051: Determine whether the shortest vertical distance to the target is less than or equal to the preset warning distance threshold.

[0118] In this embodiment of the invention, after the ground station 1 acquires the fused flight trajectory data of the abnormal UAV 4 and the flight trajectory data of the operational UAV 2, and calculates the shortest vertical distance to the target, it immediately compares the real-time quantified value with the preset warning distance threshold set by the system. Through programmed logic, it determines whether the shortest vertical distance to the target has reached the warning triggering condition, thereby determining whether the abnormal UAV 4 has posed a direct collision risk to the operational UAV 2.

[0119] It is worth mentioning that the algorithm layer in the warning stage uses image data from a single-modal high-speed camera 23 to improve the model inference speed in this stage. Only the YOLOv11-Tiny model is used, where model pruning and distillation reduce model parameters. Based on the target bounding box output by the model, the target in the point cloud data is segmented, and its centroid coordinates are obtained through bounding analysis, thus obtaining the target's flight trajectory coordinates.

[0120] Step S2052: If the distance is less than or equal to the preset warning distance threshold, then determine whether the shortest vertical distance of the target is less than or equal to the preset countermeasure distance threshold.

[0121] It should be noted that the preset countermeasure distance threshold refers to a vertical distance threshold below the warning and alarm thresholds preset for the safety of the operational UAV 2, which is the countermeasure phase of the visual dynamic scheduling strategy. Specifically, the countermeasure phase refers to the point where the ground radar 3 and airborne radar fusion perception have detected that the abnormal UAV 4 has entered the countermeasure distance range, and its continued flight will cause a collision with the operational UAV 2. The distance range of the countermeasure phase is less than 5 meters, that is, the shortest vertical distance between the flight trajectory of the abnormal UAV 4 predicted by the ground radar 3 and the flight trajectory of the abnormal UAV 4 predicted by the airborne radar and the operational UAV 2. During this countermeasure phase, the parameters of the airborne radar and high-speed camera 23 are set, with the lidar 22 undergoing 1 / 2 (adjustable) downsampling, the camera set to 240fps and 480p resolution, and the inference speed compressed to <30ms.

[0122] In this embodiment of the invention, after determining that the shortest vertical distance of the target is less than or equal to a preset alarm distance threshold, the ground station 1 needs to further determine whether the shortest vertical distance of the target is less than or equal to the preset alarm distance threshold.

[0123] Step S2053: If the distance exceeds the preset countermeasures threshold, a warning message is issued to the abnormal drone 4.

[0124] It should be noted that the warning message refers to the warning instruction issued by the system to the abnormal drone 4 after determining that there is a direct collision risk. The warning instruction includes airspace intrusion warning and avoidance requirements.

[0125] In this embodiment of the invention, when the ground station 1 determines that the shortest vertical distance to the target is less than or equal to a preset alarm distance threshold and greater than a preset countermeasure distance threshold, it immediately generates a standardized warning command and sends it synchronously to the abnormal drone 4 via a wireless communication link. At the same time, it controls the onboard alarm device of the operation drone 2 to issue a warning signal. The warning information includes airspace encroachment prompts and avoidance instructions, reminding the operator of the abnormal drone 4 to adjust the flight trajectory in time and move away from the airspace of the operation drone 2. If the abnormal drone 4 does not respond to the warning, the system will trigger subsequent higher-level countermeasure operations.

[0126] In this embodiment of the invention, after the ground station 1 determines that the shortest vertical distance of the target is greater than the preset alarm distance threshold, it does not initiate the alarm and warning process, but continues to compare the shortest vertical distance of the target with the preset countermeasure distance threshold of the system, and determines whether the vertical distance between the two aircraft reaches the countermeasure triggering condition through programmed logic, thereby determining whether the abnormal UAV 4 has formed an imminent collision risk.

[0127] In step S2054, if the distance is less than or equal to the preset countermeasure distance threshold, the corresponding countermeasure operation is performed according to the drone type of the abnormal drone 4, and the operation drone 2 is controlled to perform conflict avoidance processing.

[0128] It should be noted that countermeasures refer to the countermeasure phase of the visual dynamic scheduling strategy, and the pre-set countermeasure schemes that correspond one-to-one with different types of drones.

[0129] It is worth mentioning that, at the algorithm level of the countermeasure phase, based on the target identified in the warning phase, a tracking algorithm will be used to track the target in the countermeasure phase. Using the target identified by YOLOv11-Tiny as the initial frame, Kalman filter prediction + Hungarian algorithm matching (Meanshift algorithm, ByteTrack, etc. can be selected) is used for target tracking. The target point cloud in the segmented point cloud data of the tracked bounding box is then analyzed to obtain its centroid coordinates, and the target's flight trajectory coordinates are obtained.

[0130] In this embodiment of the invention, after determining that the shortest vertical distance to the target meets the countermeasure triggering conditions, the ground station 1 quickly retrieves the abnormal UAV 4 type information identified and stored in the early stage, matches the countermeasure strategy corresponding to the type preset by the system, and starts the dedicated countermeasure equipment to perform targeted countermeasure intervention. At the same time, it simultaneously sends obstacle avoidance control commands to the operation UAV 2, controls the operation UAV 2 to quickly adjust flight parameters such as flight altitude and flight direction, and changes the flight trajectory, so as to realize the simultaneous advancement of countermeasure operation and obstacle avoidance processing, quickly isolate the abnormal UAV 4, and completely eliminate the risk of imminent collision.

[0131] Step S206: When the abnormal drone 4 enters the warning distance range and the drone type of the abnormal drone 4 is a micro drone, the abnormal drone 4 is countered by ground countermeasure equipment.

[0132] It should be noted that ground-based countermeasures equipment refers to specialized countermeasures equipment deployed on the ground surrounding the operational airspace.

[0133] In this embodiment of the invention, when the abnormal drone 4 is identified as a micro drone, or when the target size measured by the airborne radar is less than 20cm, when the abnormal drone 4 enters the warning distance range, the ground countermeasure equipment performs countermeasure operations, such as radio frequency communication interference, net capture and interception, etc.

[0134] Step S207: When the abnormal drone 4 enters the warning distance range and the drone type of the abnormal drone 4 is a light drone, the abnormal drone 4 is countered by the ground countermeasure equipment.

[0135] It should be noted that the warning distance range refers to the airspace range centered on the operating drone 2, defined by the preset warning distance threshold, which is inside the warning distance range and outside the countermeasure distance range.

[0136] In this embodiment of the invention, when the abnormal drone 4 is identified as a light drone, or when the target size measured by the airborne radar is greater than 20cm and less than 40cm, when the abnormal drone 4 enters the warning distance range, the ground countermeasure equipment performs countermeasure operations, such as radio frequency communication interference, net capture and interception, etc.

[0137] Step S208: When the drone type of the abnormal drone 4 is a small drone and the collision time between the flight trajectory data of the abnormal drone 4 and the flight trajectory data of the operation drone 2 is less than a preset time threshold, control the operation drone 2 to perform collision avoidance processing.

[0138] It should be noted that the collision time refers to the estimated time that a collision might occur if the two drones, 4 and 2, were to maintain their current flight states, based on the flight trajectory data and flight speed of the abnormal drone 4 and the operational drone 2.

[0139] The preset time threshold refers to the time threshold set by the system to determine whether the collision risk is urgent or not, which includes but is not limited to 1 second.

[0140] In this embodiment of the invention, when the abnormal drone 4 is identified as a small drone, or when the target size measured by the airborne radar is greater than 40cm, the first collision avoidance step involves the real-time position, velocity, and predicted trajectory of the abnormal drone 4 output after Kalman filter fusion. If the collision time with the operational drone 2 is less than 1 second, the RRT* obstacle avoidance algorithm is triggered. The second collision avoidance step involves the airborne countermeasure device 21, such as net interception, being triggered when the abnormal drone 4 attacks and collides again, provided the collision time is less than 1 second.

[0141] It is worth mentioning that when the abnormal drone 4 is identified as a micro drone or a light drone, and continues to attack the operation drone 2 after evading the ground countermeasures equipment, the airborne countermeasures device 21 is triggered when the collision time of the operation drone 2 is less than 1 second, and the RRT* obstacle avoidance algorithm is triggered at the same time.

[0142] This embodiment also provides a UAV conflict avoidance device with near-far cooperative perception, which is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0143] This embodiment provides a near-far cooperative sensing UAV conflict avoidance device, applied to a UAV conflict avoidance system. The system includes an operational UAV 2 and a ground positioning platform; such as... Figure 6 As shown, it includes: The data acquisition module 301 is used to collect information about the work area where the drone 2 is performing its mission, and to analyze whether there is an abnormal drone 4 in the work area information. Prediction module 302 is used to predict whether the flight trajectory of the abnormal drone 4 will enter the warning distance range of the operational drone 2 based on the analysis results. The fusion module 303 is used to extract the flight trajectory data of the abnormal UAV 4 detected by the airborne radar on the ground positioning platform and the operation UAV 2 respectively, and perform fusion processing based on the prediction results. The calculation module 304 is used to calculate the shortest vertical distance between the abnormal drone 4 and the operational drone 2 by using the fused flight trajectory data of the abnormal drone 4 and the flight trajectory data of the operational drone 2. The control module 305 is used to control the operation drone 2 to perform collision avoidance based on the shortest vertical distance to the target.

[0144] In some alternative implementations, the acquisition module 301 includes: The data acquisition unit is used to collect information about the work area where the UAV 2 is performing its mission through a ground positioning platform; The analysis unit is used to analyze the information of the work area to determine if there are any abnormal drones. The jump unit is used to jump to the step of collecting information on the work area where the UAV 2 is performing the task through the ground positioning platform if no.

[0145] In some alternative implementations, the prediction module 302 includes: The extraction unit is used to extract the flight parameters of the abnormal drone 4 and the flight parameters of the operational drone 2 when the analysis result indicates that there is an abnormal drone 4 in the operational area information. The calculation unit is used to calculate the initial vertical shortest distance between the abnormal drone 4 and the operational drone 2 using the flight parameters of the abnormal drone 4 and the operational drone 2. The first judgment unit is used to determine whether the initial vertical shortest distance is less than or equal to the preset warning distance threshold and greater than the preset warning distance threshold. The early warning unit is used to determine, if so, that the flight trajectory of the abnormal drone 4 has entered the early warning distance range of the operational drone 2.

[0146] In some alternative implementations, the fusion module 303 includes: The activation unit is used to activate the airborne radar and high-speed camera 23 of the operational drone 2 when the flight trajectory of the abnormal drone 4 enters the warning range of the operational drone 2. The imaging unit is used to detect the flight parameters of the abnormal drone 4 through airborne radar and to capture image data of the drone by the high-speed camera 23. The combining unit is used to combine the flight parameters and image data of the abnormal UAV 4 detected by the airborne radar to generate close-range flight trajectory data of the abnormal UAV 4. The data extraction unit is used to extract the global flight trajectory data of the abnormal UAV 4 detected by the ground positioning platform; The fusion processing unit is used to fuse near-field flight trajectory data and global flight trajectory data.

[0147] In some optional implementations, the fusion processing unit includes: The preprocessing subunit is used to preprocess the near-field flight trajectory data and the global flight trajectory data; The alignment processing subunit is used to perform timestamp alignment processing on the preprocessed near-field flight trajectory data and the global flight trajectory data; The fusion subunit is used to fuse the time-stamp-aligned close-range flight trajectory data and the global flight trajectory data using the coordinate positions corresponding to the timestamp-aligned close-range flight trajectory data and the global flight trajectory data as the initial state of the Kalman filter. The fusion iterative method of Kalman filter is used to fuse the timestamp-aligned close-range flight trajectory data and the global flight trajectory data.

[0148] In some alternative implementations, the control module 305 includes: The second judgment unit is used to determine whether the shortest vertical distance of the target is less than or equal to the preset warning distance threshold. The third judgment unit is used to determine whether the shortest vertical distance of the target is less than or equal to the preset countermeasure distance threshold if it is less than or equal to the preset warning distance threshold. The alarm unit is used to issue a warning message to the abnormal drone 4 if the distance exceeds the preset countermeasures threshold. The collision avoidance unit is used to perform the corresponding countermeasure operation according to the drone type of the abnormal drone 4 if the distance is less than or equal to the preset countermeasure distance threshold, and to control the working drone 2 to perform collision avoidance processing.

[0149] In some alternative embodiments, the device further includes: The first countermeasure unit is used to perform countermeasures against the abnormal drone 4 through ground countermeasure equipment when the abnormal drone 4 enters the warning distance range and the type of drone 4 is a micro drone. The second countermeasure unit is used to perform countermeasures against the abnormal drone 4 through ground countermeasure equipment when the abnormal drone 4 enters the warning distance range and the type of the abnormal drone 4 is a light drone. The control unit is used to control the operational drone 2 to perform collision avoidance processing when the drone type of the abnormal drone 4 is a small drone and the collision time between the flight trajectory data of the abnormal drone 4 and the flight trajectory data of the operational drone 2 is less than a preset time threshold.

[0150] The UAV conflict avoidance device with near-far cooperative perception provided in this embodiment of the invention can execute the UAV conflict avoidance method with near-far cooperative perception provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0151] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0152] The following is a detailed reference. Figure 7 This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 401, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 402 or a program loaded from memory 408 into random access memory (RAM) 403. RAM 403 also stores various programs and data required for the operation of the electronic device. The processor 401, ROM 402, and RAM 403 are interconnected via bus 404. Input / output (I / O) interface 405 is also connected to bus 404.

[0153] Typically, the following devices can be connected to I / O interface 405: input devices 406 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 407 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 408 including, for example, magnetic tapes, hard disks, etc.; and communication devices 409. Communication device 409 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 7 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0154] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 409, or installed from a memory 408, or installed from a ROM 402. When the computer program is executed by the processor 401, it performs the functions defined in the near-far cooperative perception UAV conflict avoidance method of the present invention.

[0155] Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0156] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the near-far cooperative perception UAV conflict avoidance method shown in the above embodiments is implemented.

[0157] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for UAV conflict avoidance based on near-far cooperative perception, characterized in that, An application to a drone conflict and obstacle avoidance system, the system comprising an operational drone and a ground positioning platform; the method comprising: Collect information on the work area where the drone is performing its mission, and analyze whether there are any abnormal drones in the work area information; Based on the analysis results, predict whether the flight trajectory of the abnormal drone will enter the warning distance range of the operational drone; Based on the prediction results, the flight trajectory data of the abnormal UAVs detected by the airborne radar on the ground positioning platform and the operation UAV are extracted and fused. The shortest vertical distance between the abnormal drone and the operational drone is calculated by using the fused flight trajectory data of the abnormal drone and the operational drone. Based on the shortest vertical distance to the target, the operating drone is controlled to perform collision avoidance.

2. The method according to claim 1, characterized in that, The step of collecting information about the operational area where the drone is performing its mission and monitoring the operational area for any abnormal drones includes: The ground positioning platform collects information about the operational area where the drone is performing its mission. Analyze the information in the work area to determine if there are any abnormal drones; If not, proceed to the step of collecting information about the work area where the UAV is performing its mission through the ground positioning platform.

3. The method according to claim 1, characterized in that, The step of predicting whether the flight trajectory of the abnormal drone enters the warning distance range of the operational drone based on the analysis results includes: When the analysis results indicate the presence of an abnormal drone in the work area information, the flight parameters of the abnormal drone and the flight parameters of the work drone are extracted. Using the flight parameters of the abnormal UAV and the flight parameters of the operational UAV, calculate the initial vertical shortest distance between the abnormal UAV and the operational UAV; Determine whether the initial vertical shortest distance is less than or equal to a preset warning distance threshold and greater than a preset alert distance threshold; If so, it is determined that the flight trajectory of the abnormal drone has entered the warning distance range of the operational drone.

4. The method according to claim 1, characterized in that, The step of extracting and fusing the flight trajectory data of the abnormal UAVs detected by the ground positioning platform and the airborne radar on the operational UAV based on the prediction results includes: When the flight path of the abnormal drone enters the warning range of the operational drone, the onboard radar and high-speed camera of the operational drone are activated. The airborne radar detects the flight parameters of the abnormal drone, and the high-speed camera captures image data of the drone. By combining the flight parameters detected by the airborne radar and the image data, close-range flight trajectory data of the abnormal drone is generated; Extract the global flight trajectory data of the abnormal UAV detected by the ground positioning platform; The close-range flight trajectory data and the global flight trajectory data are fused together.

5. The method according to claim 4, characterized in that, The process of fusing the near-field flight trajectory data and the global flight trajectory data includes: The near-field flight trajectory data and the global flight trajectory data are preprocessed; The preprocessed near-field flight trajectory data and global flight trajectory data are timestamped and aligned. Using the coordinate positions corresponding to the timestamp-aligned close-range flight trajectory data and the global flight trajectory data as the initial state of the Kalman filter, the timestamp-aligned close-range flight trajectory data and the global flight trajectory data are fused using the Kalman filter fusion iterative method.

6. The method according to claim 1, characterized in that, The step of controlling the operational drone to perform collision avoidance based on the shortest vertical distance to the target includes: Determine whether the shortest vertical distance to the target is less than or equal to a preset warning distance threshold; If the distance is less than or equal to the preset warning distance threshold, then determine whether the shortest vertical distance to the target is less than or equal to the preset countermeasure distance threshold. If the distance exceeds the preset countermeasures threshold, a warning message is issued to the abnormal drone. If the distance is less than or equal to the preset countermeasure distance threshold, then the corresponding countermeasure operation is performed according to the type of the abnormal drone, and the operation drone is controlled to perform conflict avoidance processing.

7. The method according to claim 6, characterized in that, The method further includes: When the abnormal drone enters the warning distance range and the type of the abnormal drone is a micro drone, countermeasures are performed on the abnormal drone by ground-based countermeasure equipment. When the abnormal drone enters the warning distance range and the type of the abnormal drone is a light drone, the ground countermeasure equipment performs a countermeasure operation on the abnormal drone. When the abnormal drone is a small drone and the collision time between the flight trajectory data of the abnormal drone and the flight trajectory data of the operational drone is less than a preset time threshold, the operational drone is controlled to perform collision avoidance processing.

8. A UAV conflict avoidance device with near-far cooperative sensing, characterized in that, An application in a drone conflict and obstacle avoidance system, the system comprising an operational drone and a ground positioning platform; the device comprising: The data acquisition module is used to collect information about the work area where the drone is performing its mission, and to analyze whether there are any abnormal drones in the work area information. The prediction module is used to predict, based on the analysis results, whether the flight trajectory of the abnormal drone will enter the warning distance range of the operational drone; The fusion module is used to extract and fuse the flight trajectory data of the abnormal UAVs detected by the ground positioning platform and the airborne radar on the operation UAV, respectively, based on the prediction results. The calculation module is used to calculate the shortest vertical distance between the abnormal drone and the operational drone by using the fused flight trajectory data of the abnormal drone and the flight trajectory data of the operational drone. The control module is used to control the operating drone to perform collision avoidance based on the shortest vertical distance to the target.

9. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the UAV conflict avoidance method based on near-far cooperative perception as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the UAV conflict avoidance method of near-far cooperative perception as described in any one of claims 1 to 7.

Images