Unmanned aerial vehicle management method and device, electronic device, program product and storage medium

Abstract

The application relates to the technical field of unmanned aerial vehicle management and control, and provides an unmanned aerial vehicle management and control method and device, electronic equipment, a program product and a storage medium. The method comprises the following steps: generating an invasion alarm for an unmanned aerial vehicle based on space-time information of the unmanned aerial vehicle; determining a predicted trajectory section of the unmanned aerial vehicle based on historical trajectory points of the unmanned aerial vehicle; determining an invasion situation type of the unmanned aerial vehicle based on the predicted trajectory section; and managing and controlling the unmanned aerial vehicle based on the invasion situation type. The application can further analyze and identify the future dynamic subdivided behavior of the unmanned aerial vehicle on the basis of the static determination alarm of the unmanned aerial vehicle, so that the regulatory response is more targeted, and the demand for fine management and control in a complex low-altitude scene can be met.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) management technology, specifically to a UAV management method, device, electronic equipment, program product, and storage medium. Background Technology

[0002] With the rapid development of the low-altitude economy, drones are increasingly used in logistics, geographic surveying, emergency rescue, and consumer entertainment. The disorderly flight activities of drones can easily lead to conflicts over airspace resources and security threats in core areas, such as intrusion into airport airspace and unauthorized flights at major events. There is an urgent need to use technology to achieve real-time monitoring and risk warning of drone flight trajectories to ensure the safety of low-altitude operations. This has become a core requirement in the field of low-altitude airspace management.

[0003] In terms of drone management, existing solutions mostly focus on static judgment, that is, they only determine whether an intrusion has occurred based on the drone's current trajectory point and then issue an alarm. They cannot identify the dynamic and detailed behavior of drones, resulting in a lack of targeted management response and difficulty in meeting the needs of refined management in complex low-altitude scenarios. Summary of the Invention

[0004] This application provides a drone management method, device, electronic device, program product, and storage medium to solve the technical problem that existing solutions only statically determine whether an intrusion has occurred based on the drone's current trajectory point and then issue an alarm, which cannot identify the dynamic and detailed behavior of the drone, resulting in a lack of targeted management response and difficulty in meeting the needs of refined management in complex low-altitude scenarios.

[0005] In a first aspect, embodiments of this application provide a method for controlling unmanned aerial vehicles (UAVs), including: Based on the spatiotemporal information of the drone, an intrusion alarm is generated for the drone. Based on the historical trajectory points of the UAV, the predicted trajectory segment of the UAV is determined; Based on the predicted trajectory segment, the intrusion situation type of the UAV is determined; The drone is controlled based on the type of intrusion situation described.

[0006] In one embodiment, the spatiotemporal information includes the latitude, longitude, altitude, and time of the drone; generating an intrusion alarm for the drone based on the drone's spatiotemporal information includes: Based on the grid code corresponding to the latitude and longitude, the airspace where the UAV is located is filtered to obtain the potential conflict airspace of the UAV; Based on the range of the altitude and the time, the potential conflict airspace is filtered to obtain the filtered conflict airspace for the UAV. Based on the latitude and longitude, the filtered conflict airspace is further screened to obtain the final conflict airspace of the UAV. Based on the authorized flight activity information of the UAV and the type of the final conflict airspace, an intrusion alarm is generated for the UAV.

[0007] In one embodiment, the step of filtering the airspace where the UAV is located based on the grid code corresponding to the latitude and longitude to obtain the potential conflict airspace of the UAV includes: Based on the reverse index, obtain the spatial configuration information ID corresponding to the grid code; Based on the airspace configuration information corresponding to the airspace configuration information ID, the potential conflict airspace of the UAV is determined.

[0008] In one embodiment, filtering the potential conflict airspace based on the range of altitude and time to obtain the filtered conflict airspace for the UAV includes: If the altitude is within the altitude limit range of the potential conflict airspace and the time is within the effective time range of the potential conflict airspace, the potential conflict airspace is determined as the filtered conflict airspace for the UAV.

[0009] In one embodiment, the step of filtering the filtered conflict airspace based on the latitude and longitude to obtain the final conflict airspace for the UAV includes: Obtain the minimum bounding rectangle of the filtered conflict space; When the latitude and longitude are located within the minimum bounding rectangle, determine the relative relationship between the latitude and longitude and the spatial polygon enclosed by the filtered conflict spatial domain; If the location of the latitude and longitude is located inside or on the boundary of the airspace polygon, the filtered conflict airspace is determined as the final conflict airspace of the UAV.

[0010] In one embodiment, determining the predicted trajectory segment of the UAV based on its historical trajectory points includes: Based on the type of the drone and / or the speed of the drone, determine the number of targets and the target weights of the historical trajectory points; Based on the target weights, the historical trajectory points of the target number are weighted and fitted to obtain the UAV motion model; Based on the UAV motion model, the predicted trajectory points of the UAV are generated; Based on the predicted trajectory points, the predicted trajectory segment of the UAV is determined.

[0011] In one embodiment, determining the intrusion situation type of the UAV based on the predicted trajectory segment includes: The confidence level of the predicted trajectory segment is obtained based on the fitting error of the historical trajectory points, the deviation between the predicted trajectory segment and the actual trajectory segment, the density of the historical trajectory points, and the interference of the flight environment. Based on the relationship between the confidence level and the confidence threshold, effective predicted trajectory segments are obtained; Based on the effective predicted trajectory segment, the intrusion situation type of the UAV is determined.

[0012] In one embodiment, obtaining the effective predicted trajectory segment based on the relationship between the confidence level and the confidence threshold includes: If the confidence level is greater than or equal to the confidence level threshold, the predicted trajectory segment is determined to be a valid predicted trajectory segment; If the confidence level is less than the confidence threshold, a valid predicted trajectory segment is obtained based on a dual-mode progressive fallback strategy; the dual-mode progressive fallback strategy is a fallback strategy that progresses from the priority mode to the backup mode.

[0013] In one embodiment, obtaining the effective predicted trajectory segment based on the dual-mode progressive fallback strategy includes: Based on the priority mode, a new predicted trajectory segment is generated; the priority mode is a mode that uses the historical trajectory points of the UAV to correct the predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is less than the deviation threshold, then the new predicted trajectory segment is determined as a valid predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is greater than or equal to the deviation threshold, then a valid predicted trajectory segment is obtained based on the backup mode; the backup mode is a mode that uses the airspace boundary to constrain the new predicted trajectory segment.

[0014] In one embodiment, generating a new predicted trajectory segment based on a priority pattern includes: Obtain multiple historical trajectory points of the drone; Calculate the average latitude and longitude of the multiple historical trajectory points to obtain the location anchor point of the UAV; Based on the heading angles of several historical trajectory points among the multiple historical trajectory points, the position anchor points are corrected to obtain corrected anchor points; Based on the corrected anchor points, a new predicted trajectory segment is generated.

[0015] In one embodiment, obtaining the effective predicted trajectory segment based on the backup mode includes: Retrieve the airspace boundaries from the airspace configuration information; Based on the aforementioned spatial boundary, a trajectory constraint model is established; Based on the trajectory constraint model, the predicted trajectory points that deviate from the airspace boundary in the new predicted trajectory segment are filtered to obtain the filtered trajectory segment. If the probability that the filtered trajectory segment exceeds the spatial domain is less than or equal to the probability threshold, the filtered trajectory segment is determined as a valid predicted trajectory segment.

[0016] In one embodiment, determining the intrusion situation type of the UAV based on the effective predicted trajectory segment includes: Based on the effective predicted trajectory segment, the distance change rate, velocity vector direction, and dwell time of the UAV in the final conflict airspace are determined; Based on the distance change rate, the velocity vector direction, the dwell time, the final conflict airspace type, and the UAV type, the intrusion situation type of the UAV is determined.

[0017] In one embodiment, the rate of change of distance is obtained based on the following method: Obtain the first distance between the current trajectory point of the UAV and the center of the final conflict airspace, and obtain the second distance between the end point of the effective predicted trajectory segment and the center of the final conflict airspace; Calculate the difference between the second distance and the first distance to obtain the distance change value; The distance change rate is obtained by calculating the ratio of the distance change value to the duration of the effective predicted trajectory segment.

[0018] In one embodiment, the velocity vector direction is obtained based on the following method: Obtain the heading angle of the predicted trajectory point in the effective predicted trajectory segment; The direction of the velocity vector is determined based on the mean heading angle of the predicted trajectory points.

[0019] In one embodiment, the duration of stay is obtained based on the following method: Obtain the target predicted trajectory points located within the final conflict airspace in the effective predicted trajectory segment; If the number of target predicted trajectory points is greater than or equal to a number threshold, calculate the absolute value of the distance change rate of the target predicted trajectory points; If the absolute value is less than or equal to the absolute value threshold, the time difference between the current trajectory point of the UAV and the last trajectory point in the target predicted trajectory points is calculated, and the time difference is determined as the dwell time.

[0020] Secondly, embodiments of this application provide a drone control device, comprising: An intrusion alarm generation module is used to generate an intrusion alarm for the drone based on the drone's spatiotemporal information. The predicted trajectory determination module is used to: determine the predicted trajectory segment of the UAV based on the historical trajectory points of the UAV; The situation type determination module is used to: determine the intrusion situation type of the UAV based on the predicted trajectory segment; The drone management module is used to manage the drone based on the intrusion situation type.

[0021] Thirdly, embodiments of this application provide an electronic device, including a processor and a memory storing a computer program, wherein the processor executes the computer program to implement the steps of the drone control method described in the first aspect.

[0022] Fourthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the unmanned aerial vehicle (UAV) control method described in the first aspect.

[0023] Fifthly, embodiments of this application provide a non-transitory computer-readable storage medium, including a computer program, which, when executed by a processor, implements the steps of the unmanned aerial vehicle (UAV) control method described in the first aspect.

[0024] The drone management method, device, electronic equipment, program product, and storage medium provided in this application generate intrusion alarms for drones based on their spatiotemporal information, determine predicted trajectory segments based on historical trajectory points, determine the intrusion situation type based on the predicted trajectory segments, and manage the drone based on the intrusion situation type. In this application, the drone is first statically determined to have "intruded" based on its spatiotemporal information, and an intrusion alarm is generated. Then, based on the technical route of historical trajectory points → predicted trajectory segments → intrusion situation type → drone management, an upgrade from static alarm determination to dynamic behavior prediction is achieved. In summary, this application can further analyze and identify the future dynamic subdivisions of drone behavior based on static alarm determination, making regulatory responses more targeted and meeting the needs of refined management in complex low-altitude scenarios. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1This is one of the flowcharts illustrating the drone control method provided in the embodiments of this application; Figure 2 This is a second schematic flowchart of the drone control method provided in the embodiments of this application; Figure 3 This is the third flowchart illustrating the drone control method provided in this application embodiment; Figure 4 This is the fourth flowchart illustrating the drone control method provided in the embodiments of this application; Figure 5 This is the fifth flowchart illustrating the drone control method provided in the embodiments of this application; Figure 6 This is the sixth flowchart illustrating the drone control method provided in this application embodiment; Figure 7 This is a schematic diagram of the structure of the drone control device provided in the embodiments of this application; Figure 8 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0028] Figure 1 This is one of the flowcharts illustrating the drone control method provided in this application. (Refer to...) Figure 1 This application provides a method for controlling unmanned aerial vehicles (UAVs), which may include: Step 101: Generate an intrusion alarm for the drone based on the drone's spatiotemporal information; Step 102: Based on the historical trajectory points of the UAV, determine the predicted trajectory segment of the UAV; Step 103: Determine the type of UAV intrusion situation based on the predicted trajectory segment; Step 104: Control the drone based on the type of intrusion situation.

[0029] In step 101, based on the spatiotemporal information of the UAV, it can be determined which airspaces the current trajectory point of the UAV conflicts with, thereby generating an intrusion alarm for the conflict area.

[0030] In step 102, the future trajectory of the UAV can be predicted based on the historical trajectory points of the UAV and their corresponding weights, thereby generating a predicted trajectory segment.

[0031] In steps 103 to 104, the intrusion situation type of the UAV may include intrusion escalation, intrusion mitigation, loitering, rapid crossing, etc., which are not limited here. In this embodiment, based on the predicted trajectory segment, the movement trend of the UAV can be analyzed to identify the future dynamic subdivision behavior of the UAV, that is, the intrusion situation category of the UAV, so as to achieve more targeted and refined monitoring and response.

[0032] The drone management method provided in this embodiment generates intrusion alarms based on the drone's spatiotemporal information, determines the drone's predicted trajectory segment based on the drone's historical trajectory points, determines the drone's intrusion situation type based on the predicted trajectory segment, and manages the drone based on the intrusion situation type. In this embodiment, it first statically determines whether the drone has "intruded" based on the drone's spatiotemporal information and generates an intrusion alarm. Then, based on the technical route of historical trajectory points → predicted trajectory segment → intrusion situation type → drone management, it upgrades from static alarm determination to dynamic behavior prediction. In summary, this embodiment can further analyze and identify the drone's future dynamic subdivided behaviors on the basis of static alarm determination, making the regulatory response more targeted and meeting the needs of refined management in complex low-altitude scenarios.

[0033] Figure 2 This is the second flowchart illustrating the drone control method provided in this application. (Refer to...) Figure 2 In one embodiment, the spatiotemporal information includes the latitude, longitude, altitude, and time of the UAV; step 101 may include: Step 201: Based on the grid coding corresponding to latitude and longitude, filter the airspace where the UAV is located to obtain the potential conflict airspace of the UAV; Step 202: Based on the altitude and time range, filter the potential conflict airspace to obtain the filtered conflict airspace for the UAV. Step 203: Based on the latitude and longitude location, filter the conflict airspace to obtain the final conflict airspace for the UAV; Step 204: Based on the authorized flight activity information of the UAV and the type of the final conflict airspace, generate an intrusion alarm for the UAV.

[0034] In step 201, the grid code corresponding to latitude and longitude can be obtained based on any coding method, and there is no limitation here; in this embodiment, the grid code corresponding to latitude and longitude can be obtained by performing 17-level two-dimensional grid coding on the latitude and longitude of the current trajectory point of the UAV based on the "Earth Space Grid Coding Rules" (GB / T40087-2021).

[0035] When the current trajectory point of a drone is located within or at the boundary of a single airspace, that single airspace is the conflict airspace of the drone. When the current trajectory point of a drone is located at the boundary of multiple airspaces, these airspaces are all the conflict airspaces of the drone.

[0036] First, the airspace can be filtered based on the grid code corresponding to the latitude and longitude of the current trajectory point of the UAV to obtain potential conflict airspace. This potential conflict airspace may have some errors, so further filtering and secondary screening are required.

[0037] In step 202, the initial screening only considers the latitude and longitude of the UAV's current trajectory point. However, the altitude and time of the UAV's current trajectory point also affect its compatibility with the airspace. Therefore, based on the range of its altitude and time, the potential conflict airspace is further filtered to obtain the filtered conflict area.

[0038] In step 203, the latitude and longitude of the current trajectory point of the UAV directly determines the topological relationship between the current trajectory point and the airspace. Therefore, the filtered conflict airspace can be screened a second time to accurately determine whether the UAV is inside, on the boundary or outside of the filtered conflict airspace, thereby selecting the final conflict airspace.

[0039] In step 204, the authorized flight activity information for the drone includes whether the drone has authorized flight activities and the authorized flight range. The final conflict airspace type can include controlled areas, defense zones, temporary airspace, etc., which are not limited here. The following are some types of intrusion alarms: 1. If the drone has authorized flight activities, but its current trajectory point is not within the authorized flight range, the generated intrusion alarm will be "deviation from flight activity"; 2. If the drone is not authorized to fly and its current trajectory point is within the controlled area, the generated intrusion alarm will be "entering the controlled area"; 3. If the drone's current trajectory point is within the defense zone, the generated intrusion alarm will be "Intrusion into the defense zone"; 4. If the drone is not authorized to fly and its current trajectory point is within the temporary airspace, the generated intrusion alarm will be "entering the temporary airspace".

[0040] Furthermore, the alarm status can be updated by comparing the intrusion alarm information of the current trajectory point with the cached intrusion alarm information of the previous trajectory point in memory, and dynamically updating the alarm status. For example: 1. If the previous trajectory point had an intrusion alarm of a certain type, but the current trajectory point does not have an intrusion alarm of that type, then the alarm status will be updated to "closed" and the storage flag will be set to "updated". 2. If both the previous trajectory point and the current trajectory point have the same type of intrusion alarm, then update the alarm status to "alarming" and set the storage flag to "updating"; 3. If there is no intrusion alarm of a certain type at the previous trajectory point, but there is an intrusion alarm of that type at the current trajectory point, a new alarm record will be generated, a unique alarm number will be assigned, the alarm status will be updated to "alarming", and the storage flag will be set to "new".

[0041] In this embodiment, the airspace is first filtered based on the grid code corresponding to the latitude and longitude of the UAV's current trajectory point. Then, the airspace is filtered again based on the altitude and time range of the UAV's current trajectory point. Finally, the topological relationship between the UAV's current trajectory point and the airspace is determined based on its position, thereby filtering out the final conflicting airspace. This is equivalent to a three-layer filtering process, which can significantly reduce the amount of invalid calculations while ensuring the filtering accuracy, achieving millisecond-level filtering. Compared with the second-level filtering of full airspace polygon verification, this effectively improves the filtering efficiency. Practice has shown that the filtering accuracy can be less than or equal to 2 meters, and the amount of invalid calculations can be reduced by more than 95%. Furthermore, by comparing historical and current alarm information in memory, the system automatically generates "new / update / closed" statuses and assigns unique numbers without searching the database, reducing database pressure and realizing intelligent management of dynamic alarm status.

[0042] In one embodiment, step 201 may include: Based on the reverse index, the airspace configuration information ID corresponding to the grid code is obtained. Based on the airspace configuration information corresponding to the airspace configuration information ID, the potential conflict airspace of the UAV is determined.

[0043] All airspace configuration information is preloaded from the database into memory. This airspace configuration information includes configuration information for airspaces allocated by the system and configuration information for flight airspaces applied for by the user. The configuration information for each airspace can include airspace type (controlled area, defense area, temporary airspace, other areas, etc.), airspace boundary (vertices of a latitude-longitude polygon), airspace effective time range (start and end dates of validity), airspace altitude restriction range (minimum / maximum altitude), environmental interference level, and airspace configuration version number (format: YYYYMMDDHHMMSS_number of modifications, such as 20250908153000_001, automatically generated by the database). For flight airspaces applied for by the user, the configuration information can also include authorized flight activity information, authorized flight range information, etc.

[0044] Furthermore, a unique ID is generated for the airspace configuration information corresponding to each airspace, and a memory mapping of "airspace configuration information ID - airspace configuration information" is established. The latitude and longitude of the center point of each airspace are calculated using the polygon geometric center algorithm.

[0045] Furthermore, a triple mechanism is adopted to process airspace configuration information: "timed full verification (e.g., every 5 minutes) + incremental event push (based on RabbitMQ topic exchange) + version number verification". 1. Scheduled full verification: Compare the spatial domain configuration version number in memory with that in the database. If the memory version number is lower than the database version number, then incrementally update the difference, that is, only load the spatial domain configuration information with the higher version number in the database into memory. 2. Incremental event push: When performing add / modify / delete operations on airspace configuration information, an event is triggered and carries "operation type + airspace configuration information ID + new version number + operation timestamp". After the process instance subscribes, it only updates the corresponding airspace configuration information and synchronizes the new version number to memory. 3. Version number conflict handling: If multiple nodes push update events for the same airspace at the same time, the "operation timestamp" recorded in the database shall prevail. That is, the airspace configuration information corresponding to the latest operation timestamp shall be used as the final version, and the conflict handling log shall be written to the system log.

[0046] By adopting the above triple mechanism of "timed full verification (e.g., every 5 minutes) + incremental event push (based on RabbitMQ topic exchange) + version number verification", dynamic updates can be supported and inconsistencies between the database and the spatial configuration information in memory can be avoided.

[0047] Furthermore, based on the "Earth Space Grid Coding Rules" (GB / T 40087-2021), 17 levels of two-dimensional grid coding are applied to all airspace areas. The grid side length can be less than 500 meters, and a reverse index of "two-dimensional grid code - airspace configuration information ID - intersection type" is established. This reverse index is stored in memory, specifically as a pure memory cache. Each system instance independently caches the full data of this reverse index upon startup. Since a single two-dimensional grid may intersect with one or more airspaces, the two-dimensional grid code of that grid can be associated with the airspace configuration information IDs of all airspaces that intersect with it, thus obtaining the "two-dimensional grid code". The inverse index of "Encoding-Spatial Configuration Information ID-Intersection Type" can include 100% intersection and partial intersection. 100% intersection means that the 2D grid intersects with only one spatial domain and is located inside that spatial domain. Partial intersection means that the 2D grid intersects with only one spatial domain, and part of it is located inside that spatial domain and the other part is located outside that spatial domain. Alternatively, the 2D grid may intersect with multiple spatial domains, and each part of it may be located inside each of the multiple spatial domains.

[0048] It should be noted that for each spatial domain, as long as it intersects with any two-dimensional grid at any point, regardless of the size of the intersection area, both the spatial domain and the two-dimensional grid are included in the reverse index, achieving 100% grid coverage of the spatial domain.

[0049] Based on this reverse index, the airspace configuration information ID corresponding to the two-dimensional grid code of the latitude and longitude of the current trajectory point of the UAV and the intersection type with the airspace can be obtained. Thus, based on the memory mapping of "airspace configuration information ID - airspace configuration information", the corresponding airspace configuration information can be obtained, and the airspace described in the airspace configuration information can be identified as a potential conflict airspace.

[0050] It should be noted that in this embodiment, different levels of two-dimensional grid coding, such as 16 or 18, can be used to replace the 17-level two-dimensional grid coding for one-time screening, but the mapping logic between the grid and the spatial domain and the coding rules need to be changed; in addition, geohashing can be used to replace the "Earth Spatial Grid Coding Rules" (GB / T 40087-2021) to build a reverse index, but the method of establishing the reverse index needs to be reconstructed.

[0051] This embodiment uses two-dimensional grid coding to quickly match the corresponding airspace configuration information ID from the reverse index, thereby identifying potential conflict areas and significantly reducing the number of airspaces that need to be compared, typically by more than 95%. Specifically, since a level 17 two-dimensional grid is generally covered by 1 to 2 airspaces, that is, a two-dimensional grid code generally corresponds to 1 to 2 airspace configuration information IDs. Assuming there are a total of 100 airspaces, if a two-dimensional grid code in the reverse index corresponds to 2 airspace configuration information IDs, a hash lookup is performed based on that reverse index. One hash lookup can exclude the other 98 airspaces, that is, the number of airspaces that need to be compared is reduced by 98%, and the more airspaces, the better the effect. In addition, compared with the custom grid index, which lacks a unified standard and is prone to screening omissions and is difficult to adapt across systems, the coding index method of the "Earth Space Grid Coding Rules" (GB / T40087-2021) in this embodiment is standardized and can ensure 100% grid coverage of airspaces. Therefore, it can effectively avoid screening omissions and can be adapted across systems, thereby meeting the needs of real-time monitoring.

[0052] In one embodiment, step 202 may include: If the altitude is within the altitude limit of the potential conflict airspace and the time is within the effective time range of the potential conflict airspace, the potential conflict airspace will be identified as the filtered conflict airspace for the UAV.

[0053] For each potential conflict airspace, its airspace configuration information includes its corresponding airspace effective time range and airspace altitude restriction range. Only when the altitude of the UAV's current trajectory point is within the altitude restriction range of the potential conflict airspace (e.g., higher than its minimum altitude and / or lower than its maximum altitude), and the time of the UAV's current trajectory point is within the effective time range of the potential conflict airspace (e.g., between its start and end dates), can the potential conflict airspace be determined as the UAV's filtered conflict airspace. Airspaces that clearly do not meet the above conditions are excluded from the potential conflict airspaces, thus obtaining the UAV's filtered conflict airspace.

[0054] This embodiment, based on the initial airspace screening using the latitude and longitude of the UAV's current trajectory point, further filters the airspace based on its altitude and time, thereby obtaining a more accurate conflict airspace.

[0055] In one embodiment, step 203 may include: Obtain the minimum bounding rectangle of the filtered conflict airspace. If the latitude and longitude position is within the minimum bounding rectangle, determine the relative relationship between the latitude and longitude position and the airspace polygon enclosed by the filtered conflict airspace. If the latitude and longitude position is inside or on the boundary of the airspace polygon, determine the filtered conflict airspace as the final conflict airspace of the UAV.

[0056] It should be noted that when obtaining the potential conflict airspace of the UAV based on the inverse index "2D grid code - airspace configuration information ID - intersection type", the intersection type between the potential conflict airspace and the 2D grid containing the current trajectory point of the UAV is also obtained simultaneously. If the intersection type is 100% intersection, it means that the 2D grid intersects only with a single potential conflict airspace and is located inside that potential conflict airspace. If the potential conflict airspace is also the filtered conflict airspace of the UAV, it means that the current trajectory point inside the 2D grid must be located inside the filtered conflict airspace, meaning it can be directly... The filtered conflict airspace is determined as the final conflict airspace for the UAV. However, when the intersection type is partial intersection, it cannot be determined whether the 2D grid intersects with one or more potential conflict airspaces, nor can the positional relationship between each part of the 2D grid and the potential conflict airspace be determined. Therefore, for the current trajectory point that may exist in any region within the 2D grid, its relative positional relationship with the potential conflict airspace, i.e., its topological relationship, cannot be determined. Thus, for the filtered conflict airspace obtained from the potential conflict airspace, it is necessary to further determine its topological relationship with the current trajectory point. Specifically: First, calculate the relative relationship between the minimum bounding rectangle of the filtered conflict airspace and the latitude and longitude of the current trajectory point. If the latitude and longitude of the current trajectory point are outside the minimum bounding rectangle, the filtered conflict airspace is directly excluded to reduce the amount of calculation. If the latitude and longitude of the current trajectory point are inside the minimum bounding rectangle, the topological relationship rules of the DE-9IM (Dimensionally Extended-9 Intersection Model) matrix are further used to determine the relative relationship between the latitude and longitude of the current trajectory point and the airspace polygon enclosed by the filtered conflict airspace. If the latitude and longitude of the current trajectory point are inside or on the boundary of the airspace polygon, the filtered conflict airspace is determined as the final conflict airspace of the UAV. Based on the above method, the final conflict airspace can be further screened from the filtered conflict airspace.

[0057] It should be noted that when determining the relative relationship between the latitude and longitude position of the current trajectory point and the spatial polygon based on the DE-9IM matrix, the vertex order can be checked for consistency to ensure that the vertices of the spatial polygon are uniformly sorted in clockwise or counterclockwise order, thus avoiding judgment errors caused by disordered vertex order.

[0058] This embodiment, based on the airspace screening performed on the latitude and longitude of the UAV's current trajectory point and the airspace filtering performed on the altitude and time of the UAV's current trajectory point, further determines the topological relationship between the current trajectory point and the airspace based on the outer envelope of the airspace and the geometric verification of the DE-9IM matrix, thereby obtaining a more accurate conflict area and avoiding the error caused by the size of the two-dimensional grid itself, which would cause the airspace where the UAV flies along the outer edge to be misjudged as a conflict airspace.

[0059] Figure 3 This is the third flowchart illustrating the drone control method provided in this application. (Refer to...) Figure 3 In one embodiment, step 102 may include: Step 301: Based on the type and / or speed of the UAV, determine the number of targets and the target weights of historical trajectory points; Step 302: Based on the target weights, perform weighted fitting on the historical trajectory points of the target quantity to obtain the UAV motion model; Step 303: Based on the UAV motion model, generate the predicted trajectory points of the UAV; Step 304: Determine the predicted trajectory segment of the UAV based on the predicted trajectory points.

[0060] Drones can be categorized by their purpose, such as industrial mapping drones, logistics drones, consumer aerial photography drones, agricultural drones, inspection drones, and emergency rescue drones. Different types of drones generally have significantly different speeds, meaning there is a correlation between drone type and speed. Therefore, based on drone type, speed, or a combination of both, the number of targets and target weights for historical trajectory points can be determined as follows: 1. High-speed drones (instantaneous speed ≥ 5m / s, such as industrial surveying and logistics drones): (1) Obtain the 20 most recent historical trajectory points, covering the data of the last 20 seconds, with a sampling frequency of 1 time / second, and adopt a combination strategy of "exponential weighting + Kalman filtering for noise reduction": (2) Weighting: The first 10 historical trajectory points are assigned a dynamic weight of 60%. Specifically, the 10th historical trajectory point is used as the baseline, and its basic weight is determined by the weighting of the first 10 historical trajectory points. The weight of the 9th historical trajectory point The weight of the 8th historical trajectory point The weights of the first 10 historical trajectory points are accumulated to 60%; the 11th to 20th historical trajectory points are assigned a fixed weight of 40%. Specifically, the weight of each historical trajectory point is... ; (3) Noise reduction: Kalman filtering is applied to these historical trajectory points to eliminate position fluctuations caused by airflow interference; (4) Prediction accuracy: Predict five future trajectory points, one predicted trajectory point per second, to obtain a five-second predicted trajectory segment, with a positioning error ≤ 3 meters.

[0061] 2. Low-speed drones (instantaneous speed <1m / s, such as consumer-grade aerial photography and agricultural drones): (1) Obtain the 40 most recent historical trajectory points, covering the data of the last 40 seconds, with a sampling frequency of 1 time / second, and adopt a combination strategy of "linear weighting + sliding window smoothing": (2) Weighting: The first 10 historical trajectory points are assigned a fixed weight of 40%. Specifically, the weight of each historical trajectory point is... The 11th to 40th historical trajectory points are assigned a fixed weight of 60%. Specifically, the weight of each historical trajectory point is... ; (3) Smoothing: A sliding window is used for these historical trajectory points. The window size is 5 trajectory points. The average latitude and longitude of the trajectory points in the window is taken as the correction value of the center trajectory point in the window to suppress the slight positional shift when hovering. (4) Prediction accuracy: Predict five future trajectory points, one predicted trajectory point per second, to obtain a five-second predicted trajectory segment, with a fluctuation error ≤ 0.5 meters.

[0062] 3. Medium-speed drones (1m / s ≤ instantaneous speed < 5m / s, such as inspection and emergency rescue drones): (1) Obtain the most recent 30 historical trajectory points, covering the data of the most recent 30 seconds, with a sampling frequency of 1 time / second. The combination strategy of "velocity gradient weighting + Kalman filter denoising" is adopted. In addition, when the instantaneous speed is close to 1m / s, the Kalman filter denoising can be switched to sliding window smoothing to adapt to low-speed stability characteristics, thereby taking into account the prediction requirements of high-speed maneuverability and low-speed stability.

[0063] (2) Weight allocation: The proportion of dynamic weight and fixed weight is calculated according to the following formula, wherein the proportion of dynamic weight shall not exceed 60%, and the proportion of fixed weight shall not be less than 40%: ① Dynamic weighting percentage ; ② Fixed weight percentage .

[0064] The specific details are shown in the table below: Table 1. Example of weight allocation

[0065] (3) Noise reduction or smoothing is similar to that described above and will not be repeated here; (4) Prediction accuracy: Predict five future trajectory points, one predicted trajectory point per second, to obtain a five-second predicted trajectory segment. The combined error of positioning error and fluctuation error is ≤1.5 meters. Furthermore, when the instantaneous speed is close to 5 m / s, the combined error is ≤2 meters, and when the instantaneous speed is close to 1 m / s, the combined error is ≤1 meter.

[0066] This embodiment improves the weighted least squares method based on the type and / or speed of the drone and the requirements of low-altitude drone surveillance scenarios. It performs segmented weighting and noise reduction smoothing on the historical trajectory points of the drone, realizing scene-specific processing of the historical trajectory points. Based on the weighted fitting of these historical trajectory points, a drone motion model can be obtained. Then, based on the motion model, predicted trajectory points for future moments can be generated, thereby determining higher-precision predicted trajectory segments in subdivided scenarios.

[0067] Figure 4 This is the fourth flowchart illustrating the drone control method provided in this application. (Refer to...) Figure 4 In one embodiment, step 103 may include: Step 401: Based on the fitting error of historical trajectory points, the deviation between the predicted trajectory segment and the actual trajectory segment, the density of historical trajectory points, and flight environment interference, obtain the confidence level of the predicted trajectory segment. Step 402: Based on the relationship between confidence level and confidence threshold, obtain the effective predicted trajectory segments; Step 403: Based on the effective predicted trajectory segment, determine the intrusion situation type of the UAV.

[0068] In step 401, the confidence level of the predicted trajectory segment can be determined by combining the fitting error of historical trajectory points, the deviation between the predicted trajectory segment and the actual trajectory segment, the density of historical trajectory points, and flight environment interference in any way; no limitation is made here. In this embodiment, the confidence level of the predicted trajectory segment can be obtained by assigning scores to the fitting error of historical trajectory points, the deviation between the predicted trajectory segment and the actual trajectory segment, the density of historical trajectory points, and flight environment interference, and then performing a weighted sum. The specific calculation formula is as follows: Confidence score = Fitting error score of historical trajectory points × 50% + Deviation score between predicted trajectory segment and actual trajectory segment × 30% + Density score of historical trajectory points × 10% + Flight environment interference score × 10%; in: 1. Fitting error score of historical trajectory points: The fitting error is measured by the root mean square error. The calculation logic is that the maximum score is 50 points. If the score exceeds the threshold, points will be deducted according to the rules.

[0069] The drone type compatibility details are as follows: For high-speed drones (instantaneous speed ≥ 5m / s), a fitting error ≤ 2 meters is required to obtain full marks. 10 points will be deducted for every meter the fitting error exceeds. For low-speed drones (instantaneous speed <1m / s), a fitting error ≤0.3m is required to obtain full marks. 10 points will be deducted for every 0.1m deviation in fitting error. For medium-speed drones (1m / s ≤ instantaneous speed < 5m / s), the threshold is set by linear interpolation of the instantaneous speed. For example, when the instantaneous speed is 3m / s, the threshold is set to 1.2 meters. To get full marks, the fitting error must be ≤ 1.2 meters. 10 points will be deducted for every 0.2 meters of fitting error.

[0070] 2. Deviation score between predicted trajectory segment and actual trajectory segment: The calculation logic is a maximum score of 30 points. If the deviation exceeds the threshold, points will be deducted according to the rules.

[0071] The drone type compatibility details are unified as follows: Calculate the average deviation between the 5 predicted trajectory points and the actual trajectory points on the predicted trajectory segment. If the average deviation is ≤3 meters, full marks can be obtained. For every meter the average deviation exceeds, 5 points are deducted. If there are fewer than 5 actual trajectory points, and it is impossible to compare the 5 predicted trajectory points with the actual trajectory points, the deduction will be doubled according to the rule of "deviation value × 2". That is, the original deviation is compared with the threshold for deduction, thereby strengthening the requirement for data sufficiency.

[0072] 3. Density score of historical trajectory points: The calculation logic is out of 10 points, and the score is determined according to the actual sampling frequency of historical trajectory points.

[0073] The drone type compatibility details are unified as follows: A perfect score of 10 points is awarded if the actual sampling frequency is ≥ 1 time / second. 0.5 times / second ≤ actual sampling frequency < 1 time / second, get 5 points; If the actual sampling frequency is less than 0.5 times / second, a score of 0 is obtained. At the same time, the "data completion mechanism" is triggered to interpolate based on the flight trend of the first 10 historical trajectory points to supplement the missing historical trajectory points.

[0074] 4. Flight environment interference score: The calculation logic is a maximum score of 10 points, and the score is determined according to the "environmental interference level" in the airspace configuration information.

[0075] The drone type compatibility details are unified as follows: The final environmental interference level of the conflict airspace is low interference, such as open airspace and no electromagnetic obstruction, and scores 10 points. The final environmental interference level of the conflict airspace is medium interference, such as the urban edge and scenes with a small number of buildings obstructing the view, and it scores 5 points. The final environmental interference level of the conflict airspace is high interference, such as densely built-up areas and strong electromagnetic radiation scenarios, scoring 2 points. At the same time, the "trajectory smoothing enhancement" algorithm is activated to correct the predicted trajectory points with abnormal positions caused by interference through historical trajectory trends.

[0076] Taking a medium-speed UAV with an instantaneous speed of 3 m / s as an example, the confidence level of its predicted trajectory segment is calculated based on the above settings as follows: Assuming the fitting error of its historical trajectory points is 1.2 meters, the fitting error score of the historical trajectory points is 50 points, and the weight contribution is 50 × 50% = 25 points. Assuming the average deviation between the predicted trajectory points and the actual trajectory points is 2.5 meters, the deviation score between the predicted trajectory segment and the actual trajectory segment is 30 points, and the weight contribution is 30 × 30% = 9 points. Assuming the actual sampling frequency of its historical trajectory points is 1 time / second, the density score of the historical trajectory points is 10 points, and the weight contribution is 10 × 10% = 1 point. Assuming the environmental interference level of the final conflict airspace is medium interference, the flight environment interference score is 5 points, and the weight contribution is 5 × 10% = 0.5 points. Confidence level = 25 + 9 + 1 + 0.5 = 35.5 points, which is rounded to 36 points.

[0077] In step 402, the confidence threshold can be set based on actual needs, and is not limited here. In this embodiment, different confidence thresholds can be set for different types of drones. Since high-speed drones have strong maneuverability and require higher reliability, the confidence threshold for high-speed drones can be set to 65 points. Since medium-speed drones need to balance the maneuverability of high-speed drones and the stability of low-speed drones, the confidence threshold for medium-speed drones can be set to 60 points. Since low-speed drones are stable in flight, the threshold can be appropriately reduced, so the confidence threshold for low-speed drones is set to 55 points.

[0078] Furthermore, if the confidence level of the predicted trajectory segment is greater than or equal to the confidence level threshold, the predicted trajectory segment can be directly determined as a valid predicted trajectory segment. If the confidence level of the predicted trajectory segment is less than the confidence level threshold, the system automatically switches to a dual-mode progressive fallback strategy to obtain a valid predicted trajectory segment. This dual-mode progressive fallback strategy is a fallback strategy that progresses from a priority mode to a backup mode. That is, the priority mode is triggered first, and the backup mode is triggered when the priority mode fails, in order to achieve a fallback and ensure that in extreme scenarios, such as strong electromagnetic interference or data loss scenarios, predicted trajectory segments that meet regulatory requirements can still be output.

[0079] In step 403, since the effective predicted trajectory segment is either a predicted trajectory segment obtained through confidence evaluation or a predicted trajectory segment re-acquired based on the dual-mode progressive fallback strategy, it has higher accuracy, thus enabling accurate judgment of the UAV intrusion situation type based on the effective predicted trajectory segment.

[0080] This embodiment, based on determining a high-precision predicted trajectory segment in a subdivided scenario, further evaluates the reliability of the predicted trajectory segment by scoring its confidence level from four dimensions: fitting error of historical trajectory points, deviation between the predicted trajectory segment and the actual trajectory segment, density of historical trajectory points, and interference from the flight environment. If the reliability is high, the predicted trajectory segment is deemed valid; if the reliability is low, a dual-mode progressive fallback strategy is triggered to obtain a new valid predicted trajectory segment, thereby ensuring the accuracy of the predicted trajectory segment and further ensuring the accuracy of the UAV intrusion situation type judgment.

[0081] In one embodiment, obtaining the effective predicted trajectory segment based on a dual-mode progressive fallback strategy may include: Based on the priority mode, a new predicted trajectory segment is generated. If the deviation between the new predicted trajectory segment and the actual trajectory segment is less than the deviation threshold, the new predicted trajectory segment is determined as a valid predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is greater than or equal to the deviation threshold, a valid predicted trajectory segment is obtained based on the backup mode.

[0082] Among them, the priority mode is a mode that uses the historical trajectory points of the UAV to correct the predicted trajectory segment. In addition to the confidence level of the predicted trajectory segment being less than the confidence level threshold, the triggering conditions can also include conditions such as "no extreme data anomalies, such as more than 50% of historical trajectory points being missing or position fluctuations exceeding 10 meters". The backup mode is a mode that uses airspace boundaries to constrain the new predicted trajectory segment.

[0083] In addition, the deviation threshold can be set based on actual needs and is not limited here. In this embodiment, the deviation threshold can be set to 5 meters. Then, based on the priority mode, after generating a new predicted trajectory segment, if the deviation between the new predicted trajectory segment and the actual trajectory segment is less than 5 meters, the new predicted trajectory segment is determined as a valid predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is greater than or equal to 5 meters, then a valid predicted trajectory segment is obtained further based on the backup mode.

[0084] In this embodiment, a new predicted trajectory segment is first generated based on the priority mode, and the validity of the new predicted trajectory segment is evaluated by the deviation between the new predicted trajectory segment and the actual trajectory segment. If it is invalid, a valid predicted trajectory segment is obtained based on the backup mode. So, regardless of the confidence level of the original predicted trajectory segment, a valid predicted trajectory segment can be obtained under the dual-mode progressive fallback strategy, ensuring the accuracy of subsequent situation determination.

[0085] Figure 5 This is the fifth flowchart illustrating the drone control method provided in this application. (Refer to...) Figure 5 In one embodiment, generating a new predicted trajectory segment based on a priority mode may include: Step 501: Obtain multiple historical trajectory points of the drone; Step 502: Calculate the average latitude and longitude of multiple historical trajectory points to obtain the UAV's position anchor point; Step 503: Based on the heading angles of several historical trajectory points among multiple historical trajectory points, the position anchor points are corrected to obtain the corrected anchor points; Step 504: Generate new predicted trajectory segments based on the corrected anchor points.

[0086] In step 501, the number of historical trajectory segments can be set based on actual needs and is not limited here; in this embodiment, the most recent 30 historical trajectory points of the UAV can be obtained.

[0087] In step 502, the average latitude and longitude of these 30 historical trajectory points can be calculated, and the location of this average value can be determined as the location anchor point of the UAV to ensure that the predicted trajectory points do not deviate from the recent activity range of the UAV.

[0088] In step 503, some historical trajectory points can be selected from these 30 historical trajectory points, such as the first 10 historical trajectory points. Based on the heading angles of these 10 historical trajectory points, the aforementioned position anchor point is corrected, that is, the heading angle of the position anchor point is corrected.

[0089] In step 504, based on the corrected anchor point as the reference trajectory point, the trajectory points after the reference trajectory point are predicted, and new predicted trajectory points are generated, such as five new predicted trajectory points. Based on these five new predicted trajectory points, a new 5-second predicted trajectory segment can be generated.

[0090] In this embodiment, in priority mode, the prediction accuracy is quickly supplemented by historical trajectory points to ensure that the new predicted trajectory points are consistent with the actual flight trend of the UAV, and to avoid the confidence level being lower than expected due to the deviation of the predicted trajectory points.

[0091] Figure 6 This is the sixth flowchart illustrating the drone control method provided in this application. (Refer to...) Figure 6 In one embodiment, obtaining a valid predicted trajectory segment based on the backup mode may include: Step 601: Obtain the airspace boundaries from the airspace configuration information; Step 602: Establish a trajectory constraint model based on the spatial boundary; Step 603: Based on the trajectory constraint model, filter the predicted trajectory points that deviate from the airspace boundary in the new predicted trajectory segment to obtain the filtered trajectory segment. Step 604: If the probability that the filtered trajectory segment exceeds the spatial domain is less than or equal to the probability threshold, the filtered trajectory segment is determined as a valid predicted trajectory segment.

[0092] In step 601, the airspace boundary is obtained from the airspace configuration information corresponding to the final conflict airspace of the UAV, which is represented by the vertices of a latitude and longitude polygon.

[0093] In step 602, a trajectory constraint model can be established based on the vertex coordinates of the latitude and longitude polygon. This trajectory constraint model includes the boundary threshold of the final conflict airspace.

[0094] In step 603, based on the above boundary threshold, predicted trajectory points that deviate extremely from the final conflict airspace boundary in the new predicted trajectory segment generated by the priority mode can be filtered out, so that the filtered trajectory segment can be generated based on the remaining predicted trajectory points.

[0095] In step 604, when the filtered trajectory segment meets the restriction that the probability of exceeding the final conflict airspace range is less than or equal to the probability threshold, the filtered trajectory segment can be determined to be valid and identified as a valid predicted trajectory segment.

[0096] The probability threshold can be set based on actual needs and is not limited here; in this embodiment, the probability threshold can be set to 5%.

[0097] In the standby mode, this embodiment focuses on the core requirement of low-altitude surveillance, namely, monitoring the positional relationship between the drone and the controlled airspace. This prevents the predicted trajectory segment from completely deviating from the controlled airspace due to extreme interference, such as strong electromagnetic interference or loss of equipment signals, thus ensuring that subsequent situation assessments do not lose their surveillance significance.

[0098] In one embodiment, step 403 may include: Based on the effective predicted trajectory segments, the distance change rate, velocity vector direction, and dwell time of the UAV in the final conflict airspace are determined. Based on the distance change rate, velocity vector direction, dwell time, the type of final conflict airspace, and the type of UAV, the intrusion situation type of the UAV is determined.

[0099] in: 1. The rate of change of distance can be obtained in the following way: Obtain the first distance between the current trajectory point of the UAV and the center of the final conflict airspace, and obtain the second distance between the end point of the effective predicted trajectory segment and the center of the final conflict airspace. Calculate the difference between the second distance and the first distance to obtain the distance change value. Calculate the ratio of the distance change value to the duration of the effective predicted trajectory segment to obtain the distance change rate.

[0100] In the case where the effective predicted trajectory segment is generated based on 5 predicted trajectory points, as mentioned above, it is a 5-second predicted trajectory segment. That is, the duration of the effective predicted trajectory segment is 5 seconds. Therefore, the distance change rate is obtained by dividing the distance change value by 5 seconds.

[0101] When the distance change rate is negative, it means that the second distance is less than the first distance. In other words, the endpoint of the effective predicted trajectory segment is closer to the center of the final conflict airspace than the current trajectory point. Therefore, it can be determined that the UAV is close to the center of the airspace. If the distance change rate is positive, it means that the second distance is greater than the first distance. In other words, the endpoint of the effective predicted trajectory segment is farther from the center of the final conflict airspace relative to the current trajectory point, so it can be determined that the UAV is far away from the center of the airspace.

[0102] 2. The direction of the velocity vector can be obtained based on the following method: Obtain the heading angle of the predicted trajectory points in the effective predicted trajectory segment, and determine the direction of the velocity vector based on the mean heading angle of the predicted trajectory points.

[0103] For each predicted trajectory point in the effective predicted trajectory segment, calculate its direction angle pointing to the next predicted trajectory point, which is the heading angle of that predicted trajectory point. Taking 5 predicted trajectory points as an example, we can obtain 4 heading angles. Then, calculate the average of these 4 heading angles to obtain the velocity vector direction.

[0104] If the velocity vector direction points into the final conflict airspace, for example, if the angle between the velocity vector direction and the line connecting the current trajectory point of the UAV to the center of the final conflict airspace is less than 10 degrees, that is, if the velocity vector direction deviates from the center of the airspace by less than 10 degrees, it can be determined that the UAV's heading matches the intrusion direction.

[0105] 3. The duration of stay can be obtained based on the following methods: Obtain the target predicted trajectory points located within the final conflict airspace in the effective predicted trajectory segment. If the number of target predicted trajectory points is greater than or equal to the number threshold, calculate the absolute value of the distance change rate of the target predicted trajectory points. If the absolute value is less than or equal to the absolute value threshold, calculate the time difference between the current trajectory point of the UAV and the last trajectory point among the target predicted trajectory points, and determine the dwell time as the dwell time.

[0106] The quantity threshold and absolute value threshold can be set based on actual needs and are not limited here. In this embodiment, the quantity threshold can be set to 3 and the absolute value threshold can be set to 0.1 m / s. Then, for 5 predicted trajectory points in the effective predicted trajectory segment, if at least 3 predicted trajectory points are located in the final conflict airspace, these internal predicted trajectory points are taken as target predicted trajectory points. Based on the aforementioned distance change rate calculation method, the distance change rate of each target predicted trajectory point is calculated. Assuming the 1st, 3rd, and 4th predicted trajectory points out of the 5 predicted trajectory points are the target trajectory points, taking the 1st predicted trajectory point as an example, we obtain the first distance between the current trajectory point of the UAV and the center of the final conflict airspace, and the second distance between the 1st predicted trajectory point and the center of the final conflict airspace. We calculate the difference between the second distance and the first distance to obtain the distance change value. We then calculate the ratio of the distance change value to the duration between the current trajectory point and the 1st predicted trajectory point, i.e., 1 second, to obtain the distance change rate. We obtain the distance change rate of the 3rd and 4th predicted trajectory points in the same way. If the absolute value of the distance change rate of these 3 predicted trajectory points is less than or equal to 0.1 m / s, we can determine that the UAV has a loitering tendency, which is used to assist in the prediction of the "loitering" situation.

[0107] By combining the final conflict airspace type and UAV type, a closed-loop data flow can be achieved: "trajectory prediction → index calculation (distance change rate, velocity vector direction, and dwell time) + two-dimensional adaptation (final conflict airspace type and UAV type) → intrusion situation determination," which quantifies and distinguishes four types of intrusion situations, as shown in the table below: Table 2 Situation Assessment and Alarm Table

[0108] It should be noted that, in addition to the situation assessment based on the three-dimensional indicators of "distance change rate, velocity vector direction, and dwell time", other indicators, such as air pressure and acceleration, can be added to further improve scene adaptability and assessment accuracy.

[0109] In this embodiment, by combining three-dimensional indicators such as "distance change rate, velocity vector direction, and dwell time" and overlaying airspace type (controlled area / defense area / temporary airspace) and UAV type (consumer-grade / industrial-grade) to construct a multi-dimensional decision matrix, a multi-dimensional decision matrix is ​​constructed. This matrix quantifies and distinguishes four types of intrusion situations: "intrusion escalation, intrusion mitigation, loitering and lingering, and rapid crossing." This enables an upgrade from static judgment and alarm to dynamic behavior prediction, allowing the monitoring system to identify high-risk scenarios such as "accelerated intrusion" and "loitering and lingering" 3 to 5 seconds in advance, significantly improving the targeted nature of the response.

[0110] In one embodiment, after generating the predicted trajectory segment of the drone, the situation determination logic can be dynamically adjusted based on the degree of deviation between the predicted trajectory segment and the actual trajectory segment. At the same time, it supports custom thresholds to ensure the reliability of intrusion situation determination in complex scenarios.

[0111] The intrusion situation determination logic can be adjusted based on the following table: Table 3 Situation Judgment Logic Adjustment Table

[0112] Furthermore, it also supports customizing situation assessment thresholds through the management platform at two levels: "airspace subdivision type + UAV purpose." These thresholds are bound to airspace configuration information and are loaded into memory synchronously upon system startup. Specifically: 1. Airspace subdivision type dimension: Control areas can be subdivided into airport control areas, military control areas, urban CBD control areas, etc. Defense areas can be subdivided into oil depot defense areas, power plant defense areas, sports venue defense areas, etc. Each type of subdivided airspace can be independently configured with absolute value thresholds for distance change rate and dwell time thresholds. For example, the absolute value threshold for distance change rate for "intrusion escalation" in airport control areas can be set to 0.6 m / s, which is stricter than the 0.8 m / s threshold for ordinary control areas. 2. Drone Application Dimension: The dwell time threshold for "rapid passage" of logistics drones can be relaxed to 15 seconds, and the dwell time threshold for "loitering and lingering" of agricultural drones as industrial-grade drones can be relaxed to 90 seconds. All situation thresholds for consumer-grade entertainment drones are configured "strictly". For example, the distance change rate threshold for "intrusion intensification" can be set to 0.4m / s. 3. Activation Mode: In emergency scenarios, such as sudden security tasks, select "Online Activation", which means that the configuration will be synchronized to all nodes within 10 seconds. In regular scenarios, select "Scheduled Activation", which means that the configuration will be updated uniformly at 2:00 AM every day to avoid interference during peak business hours.

[0113] In one embodiment, when a valid predicted trajectory segment indicates that "the high-risk situation will escalate" and the confidence level is ≥80 points, a "situation escalation warning" can be triggered 1 to 2 seconds in advance, reserving a time window for regulatory action, as detailed below: 1. Triggering logic: The effective predicted trajectory segment meets the "intrusion intensification" condition twice in a row, and the effective predicted trajectory segment shows "the distance to the airspace center will be shortened by ≥10 meters within 5 seconds"; This triggering logic is based on the cruise speed of low-altitude UAVs from 1m / s to 5m / s, and a 15-meter buffer distance is reserved for handling. 2. Warning output content: includes "expected duration of escalation of intrusion, expected coordinates of the core intrusion area, and recommended countermeasures", such as "expected escalation of intrusion into the airport control area runway in 10 seconds, recommended to initiate directional electromagnetic interference"; 3. Push priority: Higher than regular alarms, directly pushed to the terminal of the regional supervisor, and simultaneously linked to on-site audible and visual alarm devices, such as the boundary alarm of the controlled area.

[0114] In one embodiment, the flight profile of the drone can be distributed and cached, specifically: Using a unique identifier for the drone, such as the device's SN (Serial Number), as a distributed lock ensures that when multiple nodes process data from different drones simultaneously, the trajectory information of the same drone is only read and written by a single node, thus avoiding concurrent conflicts. Furthermore, the cached flight information can include the time, longitude, latitude, relative ground altitude, altitude, instantaneous speed, etc. of the current trajectory point of the drone, the 300 most recent historical trajectory points reported by the drone in the last 5 minutes (if less than 300, all historical trajectory points are stored), the alarm list generated by the previous trajectory point (including intrusion alarm type, alarm status, alarm situation type, alarm number, etc.), and a list of all online drones (indexed by the drone's unique identifier). Furthermore, for each new trajectory point received, the corresponding drone's cached data is updated in real time, and historical trajectory points are retained for subsequent situational analysis.

[0115] Existing solutions mostly rely on single-machine processing of trajectory data and lack distributed caching and concurrency control mechanisms. When faced with a large number of drones reporting data simultaneously, data read / write conflicts and increased decision latency are likely to occur.

[0116] This embodiment uses the UAV serial number as the core and can be combined with Redis Red Lock to achieve distributed concurrent control. The lock timeout time is linked to the data reporting cycle and supports cross-cluster node coordination. It caches real-time trajectory points, 300 historical trajectory points and alarm lists to solve the problem of concurrent access conflicts between multiple nodes. The latency of single UAV trajectory data processing is controlled at the millisecond level.

[0117] In one embodiment, the cached list of online drones can be scanned periodically, for example, every 10 seconds. For drones that exceed a timeout threshold, such as failing to report trajectory points for more than 2 minutes, all their alarm statuses are automatically set to "off," and the storage flag is set to "updated." The scanning cycle and timeout threshold can be configured based on the drone type (e.g., consumer or industrial). For industrial drones, the timeout threshold can be set to 5 minutes to reduce accidental shutdowns in high-frequency operation scenarios.

[0118] This embodiment periodically scans the cached list of online drones, supports configuring timeout thresholds by drone type, and automatically disables alarms for drones that fail to report trajectory points within the timeout period, thereby reducing invalid resource usage and improving the system's level of automated management.

[0119] In one embodiment, the processed alarm information can also be transmitted to the alarm storage system, and differentiated persistent storage can be performed based on the storage flag (new / updated). Specifically: For the "new" flag, complete alarm information (including the drone's initial location, alarm type, intrusion situation type, etc.) will be written into the database; For the "Update" flag, only the changed fields such as alarm status, drone current location, and intrusion situation type corresponding to the alarm number in the database are updated, and the complete alarm information is not stored repeatedly. Furthermore, an asynchronous batch update mechanism is adopted, such as a "time window (e.g., 0.5 seconds) + alarm type merging" mechanism. For the reported data of the detection, the data reporting frequency is high, and multiple data may be reported per second. When the same type of alarm of the same drone may be triggered multiple times within 0.5 seconds, only the latest record is retained. For example, merging two "entry into controlled area" alarms into one update record can reduce the number of database writes by 70%. Furthermore, a hot and cold data separation mechanism is adopted. Real-time alarm information is stored in a MySQL database that supports transaction processing. Closed alarm information in historical alarm information will be automatically copied and archived to the Doris database after a preset time, such as 10 minutes. This database supports large-scale data storage and efficient querying. The response time for multi-dimensional conditional queries based on drones, alarm types, and time ranges can be less than or equal to 100ms.

[0120] In this embodiment, alarm information is stored differently according to "new / update" and an asynchronous batch update mechanism is adopted, which greatly reduces the number of database operations and improves data processing efficiency. Real-time alarm information and historical alarm information are stored in relational database and analytical database respectively to ensure the efficiency and transaction characteristics of real-time alarm information, as well as the large-capacity storage and subsequent analysis of historical alarm information. A scheduled task copies closed alarm information from historical alarm information to the analytical database and then deletes it from the relational database to ensure that the number of alarm information in the relational database is small, thereby improving the efficiency of data entry and update.

[0121] The drone control device provided in the embodiments of this application is described below. The drone control device described below and the drone control method described above can be referred to in correspondence.

[0122] Figure 7 This is a schematic diagram of the unmanned aerial vehicle (UAV) control device provided in an embodiment of this application. (Refer to...) Figure 7 This application provides a drone control device, which may include: The intrusion alarm generation module 701 is used to: generate an intrusion alarm for the drone based on the spatiotemporal information of the drone; The predicted trajectory determination module 702 is used to: determine the predicted trajectory segment of the UAV based on the historical trajectory points of the UAV; The situation type determination module 703 is used to: determine the intrusion situation type of the UAV based on the predicted trajectory segment; The drone management module 704 is used to manage the drone based on the intrusion situation type.

[0123] The drone management device provided in this embodiment generates an intrusion alarm based on the drone's spatiotemporal information, determines the drone's predicted trajectory segment based on the drone's historical trajectory points, determines the drone's intrusion situation type based on the predicted trajectory segment, and manages the drone based on the intrusion situation type. In this embodiment, it first statically determines whether the drone "intrudes" based on the drone's spatiotemporal information and generates an intrusion alarm, then upgrades from static alarm determination to dynamic behavior prediction based on the technical route of historical trajectory points → predicted trajectory segment → intrusion situation type → drone management. In summary, this embodiment can further analyze and identify the drone's future dynamic behaviors based on the static alarm determination, making the regulatory response more targeted and meeting the needs of refined management in complex low-altitude scenarios.

[0124] In one embodiment, the spatiotemporal information includes the latitude, longitude, altitude, and time of the UAV; the intrusion alarm generation module 701 is specifically used for: Based on the grid code corresponding to the latitude and longitude, the airspace where the UAV is located is filtered to obtain the potential conflict airspace of the UAV; Based on the range of the altitude and the time, the potential conflict airspace is filtered to obtain the filtered conflict airspace for the UAV. Based on the latitude and longitude, the filtered conflict airspace is further screened to obtain the final conflict airspace of the UAV. Based on the authorized flight activity information of the UAV and the type of the final conflict airspace, an intrusion alarm is generated for the UAV.

[0125] In one embodiment, the intrusion alarm generation module 701 is specifically used for: Based on the reverse index, obtain the spatial configuration information ID corresponding to the grid code; Based on the airspace configuration information corresponding to the airspace configuration information ID, the potential conflict airspace of the UAV is determined.

[0126] In one embodiment, the intrusion alarm generation module 701 is specifically used for: If the altitude is within the altitude limit range of the potential conflict airspace and the time is within the effective time range of the potential conflict airspace, the potential conflict airspace is determined as the filtered conflict airspace for the UAV.

[0127] In one embodiment, the intrusion alarm generation module 701 is specifically used for: Obtain the minimum bounding rectangle of the filtered conflict space; When the latitude and longitude are located within the minimum bounding rectangle, determine the relative relationship between the latitude and longitude and the spatial polygon enclosed by the filtered conflict spatial domain; If the location of the latitude and longitude is located inside or on the boundary of the airspace polygon, the filtered conflict airspace is determined as the final conflict airspace of the UAV.

[0128] In one embodiment, the trajectory prediction and determination module 702 is specifically used for: Based on the type of the drone and / or the speed of the drone, determine the number of targets and the target weights of the historical trajectory points; Based on the target weights, the historical trajectory points of the target number are weighted and fitted to obtain the UAV motion model; Based on the UAV motion model, the predicted trajectory points of the UAV are generated; Based on the predicted trajectory points, the predicted trajectory segment of the UAV is determined.

[0129] In one embodiment, the situation type determination module 703 is specifically used for: The confidence level of the predicted trajectory segment is obtained based on the fitting error of the historical trajectory points, the deviation between the predicted trajectory segment and the actual trajectory segment, the density of the historical trajectory points, and the interference of the flight environment. Based on the relationship between the confidence level and the confidence threshold, effective predicted trajectory segments are obtained; Based on the effective predicted trajectory segment, the intrusion situation type of the UAV is determined.

[0130] In one embodiment, the situation type determination module 703 is specifically used for: If the confidence level is greater than or equal to the confidence level threshold, the predicted trajectory segment is determined to be a valid predicted trajectory segment; If the confidence level is less than the confidence threshold, a valid predicted trajectory segment is obtained based on a dual-mode progressive fallback strategy; the dual-mode progressive fallback strategy is a fallback strategy that progresses from the priority mode to the backup mode.

[0131] In one embodiment, the situation type determination module 703 is specifically used for: Based on the priority mode, a new predicted trajectory segment is generated; the priority mode is a mode that uses the historical trajectory points of the UAV to correct the predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is less than the deviation threshold, then the new predicted trajectory segment is determined as a valid predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is greater than or equal to the deviation threshold, then a valid predicted trajectory segment is obtained based on the backup mode; the backup mode is a mode that uses the airspace boundary to constrain the new predicted trajectory segment.

[0132] In one embodiment, the situation type determination module 703 is specifically used for: Obtain multiple historical trajectory points of the drone; Calculate the average latitude and longitude of the multiple historical trajectory points to obtain the location anchor point of the UAV; Based on the heading angles of several historical trajectory points among the multiple historical trajectory points, the position anchor points are corrected to obtain corrected anchor points; Based on the corrected anchor points, a new predicted trajectory segment is generated.

[0133] In one embodiment, the situation type determination module 703 is specifically used for: Retrieve the airspace boundaries from the airspace configuration information; Based on the aforementioned spatial boundary, a trajectory constraint model is established; Based on the trajectory constraint model, the predicted trajectory points that deviate from the airspace boundary in the new predicted trajectory segment are filtered to obtain the filtered trajectory segment. If the probability that the filtered trajectory segment exceeds the spatial domain is less than or equal to the probability threshold, the filtered trajectory segment is determined as a valid predicted trajectory segment.

[0134] In one embodiment, the situation type determination module 703 is specifically used for: Based on the effective predicted trajectory segment, the distance change rate, velocity vector direction, and dwell time of the UAV in the final conflict airspace are determined; Based on the distance change rate, the velocity vector direction, the dwell time, the final conflict airspace type, and the UAV type, the intrusion situation type of the UAV is determined.

[0135] In one embodiment, the situation type determination module 703 is specifically used for: Obtain the first distance between the current trajectory point of the UAV and the center of the final conflict airspace, and obtain the second distance between the end point of the effective predicted trajectory segment and the center of the final conflict airspace; Calculate the difference between the second distance and the first distance to obtain the distance change value; The distance change rate is obtained by calculating the ratio of the distance change value to the duration of the effective predicted trajectory segment.

[0136] In one embodiment, the situation type determination module 703 is specifically used for: Obtain the heading angle of the predicted trajectory point in the effective predicted trajectory segment; The direction of the velocity vector is determined based on the mean heading angle of the predicted trajectory points.

[0137] In one embodiment, the situation type determination module 703 is specifically used for: Obtain the target predicted trajectory points located within the final conflict airspace in the effective predicted trajectory segment; If the number of target predicted trajectory points is greater than or equal to a number threshold, calculate the absolute value of the distance change rate of the target predicted trajectory points; If the absolute value is less than or equal to the absolute value threshold, the time difference between the current trajectory point of the UAV and the last trajectory point in the target predicted trajectory points is calculated, and the time difference is determined as the dwell time.

[0138] Figure 8 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application, such as... Figure 8 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call the computer program in the memory 830 to execute the steps of the drone control method, such as: Based on the spatiotemporal information of the drone, an intrusion alarm is generated for the drone. Based on the historical trajectory points of the UAV, the predicted trajectory segment of the UAV is determined; Based on the predicted trajectory segment, the intrusion situation type of the UAV is determined; The drone is controlled based on the type of intrusion situation described.

[0139] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0140] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the steps of the drone control method provided in the above embodiments, such as including: Based on the spatiotemporal information of the drone, an intrusion alarm is generated for the drone. Based on the historical trajectory points of the UAV, the predicted trajectory segment of the UAV is determined; Based on the predicted trajectory segment, the intrusion situation type of the UAV is determined; The drone is controlled based on the type of intrusion situation described.

[0141] On the other hand, embodiments of this application also provide a non-transitory computer-readable storage medium storing a computer program thereon, the computer program being used to cause a processor to execute the steps of the drone control method provided in the above embodiments, for example including: Based on the spatiotemporal information of the drone, an intrusion alarm is generated for the drone. Based on the historical trajectory points of the UAV, the predicted trajectory segment of the UAV is determined; Based on the predicted trajectory segment, the intrusion situation type of the UAV is determined; The drone is controlled based on the type of intrusion situation described.

[0142] The non-transitory computer-readable storage medium can be any available medium or data storage device that the processor can access, including but not limited to magnetic memory (e.g., floppy disk, hard disk, magnetic tape, magneto-optical disk (MO)), optical memory (e.g., CD, DVD, BD, HVD), and semiconductor memory (e.g., ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state drive (SSD)).

[0143] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0144] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0145] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for controlling unmanned aerial vehicles (UAVs), characterized in that, include: Based on the spatiotemporal information of the drone, an intrusion alarm is generated for the drone. Based on the historical trajectory points of the UAV, the predicted trajectory segment of the UAV is determined; Based on the predicted trajectory segment, the intrusion situation type of the UAV is determined, including: The confidence level of the predicted trajectory segment is obtained based on the fitting error of the historical trajectory points, the deviation between the predicted trajectory segment and the actual trajectory segment, the density of the historical trajectory points, and the interference of the flight environment. If the confidence level is greater than or equal to the confidence level threshold, the predicted trajectory segment is determined to be a valid predicted trajectory segment; If the confidence level is less than the confidence threshold, a new predicted trajectory segment is generated based on a priority mode; the priority mode is a mode that uses the historical trajectory points of the UAV to correct the predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is less than the deviation threshold, then the new predicted trajectory segment is determined as a valid predicted trajectory segment. If the deviation between the new predicted trajectory segment and the actual trajectory segment is greater than or equal to the deviation threshold, then a valid predicted trajectory segment is obtained based on the backup mode; the backup mode is a mode that uses the airspace boundary to constrain the new predicted trajectory segment. Based on the effective predicted trajectory segment, the intrusion situation type of the UAV is determined; The drone is controlled based on the type of intrusion situation described.

2. The drone control method according to claim 1, characterized in that, The spatiotemporal information includes the latitude, longitude, altitude, and time of the drone; the generation of an intrusion alarm for the drone based on the drone's spatiotemporal information includes: Based on the grid code corresponding to the latitude and longitude, the airspace where the UAV is located is filtered to obtain the potential conflict airspace of the UAV; Based on the range of the altitude and the time, the potential conflict airspace is filtered to obtain the filtered conflict airspace for the UAV. Based on the latitude and longitude, the filtered conflict airspace is further screened to obtain the final conflict airspace of the UAV. Based on the authorized flight activity information of the UAV and the type of the final conflict airspace, an intrusion alarm is generated for the UAV.

3. The unmanned aerial vehicle (UAV) control method according to claim 2, characterized in that, The process of filtering the airspace where the UAV is located based on the grid code corresponding to the latitude and longitude to obtain the potential conflict airspace of the UAV includes: Based on the reverse index, obtain the spatial configuration information ID corresponding to the grid code; Based on the airspace configuration information corresponding to the airspace configuration information ID, the potential conflict airspace of the UAV is determined.

4. The unmanned aerial vehicle (UAV) control method according to claim 2, characterized in that, The filtering of the potential conflict airspace based on the altitude and time range to obtain the filtered conflict airspace for the UAV includes: If the altitude is within the altitude limit range of the potential conflict airspace and the time is within the effective time range of the potential conflict airspace, the potential conflict airspace is determined as the filtered conflict airspace for the UAV.

5. The unmanned aerial vehicle (UAV) control method according to claim 2, characterized in that, The process of filtering the conflict airspace based on the latitude and longitude coordinates to obtain the final conflict airspace for the UAV includes: Obtain the minimum bounding rectangle of the filtered conflict space; When the latitude and longitude are located within the minimum bounding rectangle, determine the relative relationship between the latitude and longitude and the spatial polygon enclosed by the filtered conflict spatial domain; If the location of the latitude and longitude is located inside or on the boundary of the airspace polygon, the filtered conflict airspace is determined as the final conflict airspace of the UAV.

6. The unmanned aerial vehicle (UAV) control method according to claim 1, characterized in that, Determining the predicted trajectory segment of the UAV based on its historical trajectory points includes: Based on the type of the drone and / or the speed of the drone, determine the number of targets and the target weights of the historical trajectory points; Based on the target weights, the historical trajectory points of the target number are weighted and fitted to obtain the UAV motion model; Based on the UAV motion model, the predicted trajectory points of the UAV are generated; Based on the predicted trajectory points, the predicted trajectory segment of the UAV is determined.

7. The unmanned aerial vehicle (UAV) control method according to claim 1, characterized in that, The generation of new predicted trajectory segments based on the priority mode includes: Obtain multiple historical trajectory points of the drone; Calculate the average latitude and longitude of the multiple historical trajectory points to obtain the location anchor point of the UAV; Based on the heading angles of several historical trajectory points among the multiple historical trajectory points, the position anchor points are corrected to obtain corrected anchor points; Based on the corrected anchor points, a new predicted trajectory segment is generated.

8. The unmanned aerial vehicle (UAV) control method according to claim 1, characterized in that, The process of obtaining effective predicted trajectory segments based on the backup mode includes: Retrieve the airspace boundaries from the airspace configuration information; Based on the aforementioned spatial boundary, a trajectory constraint model is established; Based on the trajectory constraint model, the predicted trajectory points that deviate from the airspace boundary in the new predicted trajectory segment are filtered to obtain the filtered trajectory segment. If the probability that the filtered trajectory segment exceeds the spatial domain is less than or equal to the probability threshold, the filtered trajectory segment is determined as a valid predicted trajectory segment.

9. The unmanned aerial vehicle (UAV) control method according to claim 1, characterized in that, The determination of the intrusion situation type of the UAV based on the effective predicted trajectory segment includes: Based on the effective predicted trajectory segment, the distance change rate, velocity vector direction, and dwell time of the UAV in the final conflict airspace are determined; Based on the distance change rate, the velocity vector direction, the dwell time, the final conflict airspace type, and the UAV type, the intrusion situation type of the UAV is determined.

10. The unmanned aerial vehicle (UAV) control method according to claim 9, characterized in that, The rate of change of distance was obtained based on the following method: Obtain the first distance between the current trajectory point of the UAV and the center of the final conflict airspace, and obtain the second distance between the end point of the effective predicted trajectory segment and the center of the final conflict airspace; Calculate the difference between the second distance and the first distance to obtain the distance change value; The distance change rate is obtained by calculating the ratio of the distance change value to the duration of the effective predicted trajectory segment.

11. The unmanned aerial vehicle (UAV) control method according to claim 9, characterized in that, The direction of the velocity vector is obtained based on the following method: Obtain the heading angle of the predicted trajectory point in the effective predicted trajectory segment; The direction of the velocity vector is determined based on the mean heading angle of the predicted trajectory points.

12. The unmanned aerial vehicle (UAV) control method according to claim 9, characterized in that, The duration of stay was obtained based on the following method: Obtain the target predicted trajectory points located within the final conflict airspace in the effective predicted trajectory segment; If the number of target predicted trajectory points is greater than or equal to a number threshold, calculate the absolute value of the distance change rate of the target predicted trajectory points; If the absolute value is less than or equal to the absolute value threshold, the time difference between the current trajectory point of the UAV and the last trajectory point in the target predicted trajectory points is calculated, and the time difference is determined as the dwell time.

13. A drone control device, characterized in that, The method for implementing the drone control method of claim 1 includes: An intrusion alarm generation module is used to generate an intrusion alarm for the drone based on the drone's spatiotemporal information. The predicted trajectory determination module is used to: determine the predicted trajectory segment of the UAV based on the historical trajectory points of the UAV; The situation type determination module is used to: determine the intrusion situation type of the UAV based on the predicted trajectory segment; The drone management module is used to manage the drone based on the intrusion situation type.

14. An electronic device comprising a processor and a memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the unmanned aerial vehicle (UAV) control method according to any one of claims 1 to 12.

15. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the unmanned aerial vehicle (UAV) control method according to any one of claims 1 to 12.

16. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the unmanned aerial vehicle (UAV) control method according to any one of claims 1 to 12.

Images