A method and system for swarm scheduling of police drones

Abstract

This application relates to the field of unmanned aerial vehicle (UAV) scheduling technology, and discloses a method and system for scheduling police UAV swarms. The method includes: acquiring police mission instructions issued by a police UAV scheduling system; parsing the intent of the police mission instructions to extract the mission intent and expected flight behavior; converting the extracted instructions into a semantic flight plan understandable by an external civil low-altitude traffic management system; submitting the semantic flight plan and receiving returned airspace information; acquiring manual intervention instructions from a ground command center; identifying the emergency intent of the manual intervention instructions; comprehensively evaluating the manual intervention instructions; and adjusting the anti-collision logic parameters within the police UAV swarm based on the comprehensive evaluation results; acquiring the behavioral change intent of UAVs within the police UAV swarm; and coordinating the UAVs within the police UAV swarm to perform cooperative avoidance or formation adjustments. This application effectively solves the problem of semantic incompatibility between internal police system instructions and external system instructions, and improves the accuracy of instruction parsing.

Description

Technical Field

[0001] This application relates to the field of drone scheduling technology, and more specifically, to a method and system for scheduling police drone swarms. Background Technology

[0002] In urban low-altitude airspace, police drone swarms are a vital force for daily patrols. Initially, the mission allocation and flight paths of these drones were determined primarily by pre-defined geographical divisions and police priority requirements, aiming to efficiently cover target areas and conserve resources. However, with the booming development of the low-altitude economy, the use of urban low-altitude airspace has become unprecedentedly busy. Besides police drones, civilian logistics drones, industrial drones, and even future manned aircraft operate in the same airspace, making the airspace environment extremely congested and complex. To effectively manage the increasingly busy low-altitude traffic, urban low-altitude management departments have begun implementing a new dynamic airspace allocation and management mechanism. This mechanism requires all low-altitude aircraft, including police drones, to submit flight plans through a new information exchange protocol and obtain flight permits in real time during specific busy periods. This means that police drone swarms can no longer plan and execute missions entirely autonomously; their flight activities must interact and coordinate in real time with the broader low-altitude traffic management system.

[0003] When the police drone dispatch system synchronizes information with the new civil low-altitude traffic management system, subtle differences in the definition and processing logic of specific data fields in the information exchange protocol lead to problems in data transmission and parsing. For example, the police system may use an internally defined task priority coding system, while the civil system expects a text description based on industry standards. Furthermore, inconsistencies may exist in the precision of geographic coordinate representation, the synchronization mechanism of timestamps, or the update frequency of aircraft status information. These seemingly minor differences, during peak low-altitude traffic periods when a large number of aircraft simultaneously request airspace permission, can significantly increase the response delay for the police drone dispatch system to obtain airspace permission requests, and may even occasionally result in request timeouts. This delay or failure prevents the police drone swarm from obtaining the latest airspace authorization in a timely manner, thus affecting the immediacy and flexibility of its missions.

[0004] Suppose an emergency occurs in the city, such as a public safety incident, requiring a rapid response and on-site reconnaissance support from a swarm of police drones. However, due to delays in obtaining airspace permits, some drones in the swarm cannot obtain the optimal flight path authorization to reach the scene immediately. These drones are forced to choose longer detour routes to avoid airspace without prior authorization, or they can only wait in place until permission is granted. This forced route adjustment or waiting directly prolongs the swarm's arrival time at the scene, thus missing a critical response window and adversely affecting police operations.

[0005] Faced with an emergency, the ground command center attempted manual intervention to buy precious time, forcing some drones to take off along suboptimal routes to bypass the waiting area. However, due to a data synchronization problem between the police dispatch system and the new civilian management mechanism, this manual instruction conflicted with the airspace information upon which the collision avoidance logic of the automatic dispatch system relied. For example, a manual instruction might direct a drone to enter a certain airspace, but the automatic collision avoidance system, based on its own (potentially outdated or incomplete) airspace information, determined that the airspace posed a potential collision risk. This inconsistency triggered the obstacle avoidance mechanisms of multiple drones, causing them to attempt evasive maneuvers while simultaneously executing manual instructions, further exacerbating formation chaos and reducing overall efficiency within the swarm.

[0006] Due to the persistent conflict between manual commands and automatic collision avoidance logic, the flight status within the swarm became extremely unstable. Some drones, upon receiving manual commands from the ground command center, attempted to accelerate towards the target area to execute emergency missions. However, other drones, due to their internal collision avoidance logic, detected danger and performed emergency hovering or evasive maneuvers. This conflicting command structure of advancing and avoiding quickly propagated within the swarm, creating a chain reaction. Ultimately, the entire swarm fell into a command deadlock, with the drones repeatedly making small-scale evasive maneuvers and adjustments in the air, unable to form an effective, unified direction of advance. This internal conflict caused the swarm to stagnate over critical areas, completely losing its ability to respond quickly to emergencies, making it impossible to carry out police missions as intended.

[0007] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this application provides a method and system for scheduling police drone swarms, aiming to solve technical problems such as increasingly busy urban low-altitude airspace, difficulties in synchronizing information between the police drone scheduling system and the civil low-altitude traffic management system, delays in obtaining airspace permits, conflicts between manual intervention commands and automatic collision avoidance logic, and mission execution obstruction caused by conflicting commands within the swarm.

[0009] Firstly, this application provides a method for scheduling police drone swarms, including: Obtain police mission instructions issued by the police drone dispatch system, perform intent analysis on the police mission instructions, and extract the mission intent and expected flight behavior; The mission intent and expected flight behavior are converted into a semantic flight plan that can be understood by an external civil low-altitude traffic management system, the semantic flight plan is submitted to the external civil low-altitude traffic management system, and airspace information returned by the external civil low-altitude traffic management system is received. Obtain human intervention instructions from the ground command center, identify the emergency intent of the human intervention instructions, and comprehensively evaluate the human intervention instructions based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and airspace information. Based on the comprehensive evaluation results, adjust the anti-collision logic parameters within the police drone cluster. The system obtains the behavioral change intentions of drones within the police drone swarm. Based on these intentions, the real-time situation of the police drone swarm, and the collision avoidance logic parameters, it coordinates the drones within the swarm to perform collaborative avoidance or formation adjustments.

[0010] This technical solution effectively addresses the scheduling challenges faced by police drone swarms in complex low-altitude airspace. By engaging in semantic interaction with external civil low-altitude traffic management systems, real-time airspace information is obtained. Combined with manual intervention commands and internal swarm situation assessment, the collision avoidance logic is dynamically adjusted, ultimately achieving collaborative avoidance and formation adjustment of the swarm drones. This ensures the timeliness and safety of police missions and overcomes the problems of delayed airspace permit acquisition and command conflicts leading to decreased swarm efficiency in existing technologies.

[0011] Furthermore, the steps for obtaining police mission instructions issued by the police drone dispatch system, performing intent parsing on the police mission instructions, and extracting the mission intent and expected flight behavior include: The predefined task semantic dictionary is used to match keywords in police task instructions to determine core intent labels and expected flight behavior parameters. The core intent labels are converted into task intents, and the expected flight behavior parameters are converted into expected flight behaviors. The predefined task semantic dictionary stores commonly used police task instructions and their corresponding core intent labels and expected flight behavior parameters.

[0012] Based on this, the steps to translate mission intent and anticipated flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system include: The task priority in the mission intent is mapped to a text description accepted by the external civil low-altitude traffic management system through the task priority encoding inside the police system; The geographic coordinates in the expected flight behavior are standardized based on the internationally accepted coordinate system. The timestamps in the expected flight behavior are synchronized based on the network time protocol or the precision time protocol in order to synchronize with the external civil low-altitude traffic management system. The mapped text description, standardized geographic coordinates, and synchronized timestamps are integrated and converted into a semantic flight plan.

[0013] In some preferred embodiments, the method further includes a step performed after receiving airspace information returned by an external civil low-altitude traffic management system: Analyze airspace information to obtain no-fly altitudes, temporary avoidance zones, and areas with high concentrations of civilian aircraft; No-fly altitudes, temporary avoidance zones, and densely populated areas of civilian aircraft are converted into dynamic airspace boundary parameters or obstacle avoidance behavior adjustment instructions that can be identified by the police drone collision avoidance system.

[0014] Furthermore, based on the priority of the emergency intent, the real-time situation of the current police drone swarm, and airspace information, the steps for comprehensively evaluating manual intervention commands include: The system obtains the current alarm level and, based on the current alarm level, the priority of the emergency intent, the real-time situation of the current police drone swarm, and the dynamic airspace boundary parameters or obstacle avoidance behavior adjustment instructions, comprehensively assesses the urgency and potential safety risks of the manual intervention instruction. The system calculates the task urgency score and the potential collision probability as the comprehensive assessment result.

[0015] Preferably, the step of adjusting the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results includes: When the mission urgency score is higher than the preset urgency threshold and the potential collision probability is lower than the preset collision threshold, the collision avoidance distance threshold within the police drone cluster is adjusted, a temporary safe flight path is generated, and real-time risk alerts are provided to the ground command center.

[0016] Based on the above, the steps for obtaining the behavioral change intentions of drones within a police drone swarm, and coordinating drones within the swarm to perform collaborative avoidance or formation adjustments based on these behavioral change intentions, the current real-time situation of the police drone swarm, and collision avoidance logic parameters, include: When a drone in a police drone swarm adjusts its flight behavior due to performing an emergency mission or due to human intervention, it acquires the intention to change its behavior and broadcasts the intention to change its behavior to other drones in the police drone swarm. Based on the behavioral change intent, the real-time situation of the current police drone swarm, and the collision avoidance logic parameters, the system calculates collaborative avoidance flight paths or formation reconfiguration instructions for neighboring drones.

[0017] As a technological improvement, nearby drones are identified through the following steps: Maintain a real-time cluster member location list. Based on the real-time cluster member location list, identify all neighboring drones within a preset range of the drone with behavioral change intentions. All police drones report their real-time location, speed, attitude, battery level, and mission status at a preset frequency. The real-time cluster member location list is updated in real time based on the real-time location, speed, attitude, battery level, and mission status.

[0018] As a further improvement, behavioral change intentions include the drone's unique identifier, new target location, estimated time of arrival, and new speed and altitude.

[0019] Secondly, this application also discloses a police drone swarm dispatching system for executing the above-mentioned police drone swarm dispatching method, the system comprising: The instruction intent parsing module is used to obtain police mission instructions issued by the police drone dispatch system, perform intent parsing on the police mission instructions, and extract the mission intent and expected flight behavior. The airspace information receiving module is used to convert mission intent and expected flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system, submit the semantic flight plan to the external civil low-altitude traffic management system, and receive airspace information returned by the external civil low-altitude traffic management system. The anti-collision logic parameter adjustment module is used to obtain manual intervention instructions from the ground command center, identify the emergency intent of the manual intervention instructions, comprehensively evaluate the manual intervention instructions based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and airspace information, and adjust the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results. The collaborative avoidance and formation adjustment module is used to obtain the behavioral change intentions of drones within the police drone cluster. Based on the behavioral change intentions, the real-time situation of the current police drone cluster, and the collision avoidance logic parameters, it coordinates the drones within the police drone cluster to perform collaborative avoidance or formation adjustment.

[0020] In summary, this application provides a method and system for scheduling police drone swarms. The method acquires police mission instructions and performs intent parsing to extract mission intent and expected flight behavior, achieving a precise understanding of the police mission. Subsequently, this information is converted into a semantic flight plan understandable by an external civil low-altitude traffic management system and submitted to obtain airspace information. This effectively solves the data transmission and parsing problems caused by differences in information exchange protocols between the police and civil systems, significantly reducing response latency and failure rate for airspace permission requests. Upon receiving manual intervention instructions from the ground command center, the method can identify emergency intents and perform a comprehensive evaluation based on the priority of the emergency intent, the real-time situation of the current police drone swarm, and airspace information. Based on the evaluation results, it dynamically adjusts the anti-collision logic parameters within the police drone swarm. This mechanism effectively overcomes the conflict between manual instructions and automatic anti-collision logic in existing technologies, avoiding internal swarm chaos and efficiency degradation. Ultimately, by acquiring the behavioral intentions of drones within the swarm, and based on these intentions, the swarm's real-time situation, and collision avoidance logic parameters, the drones are coordinated to perform cooperative avoidance or formation adjustments. This completely resolves the problem of swarms stalling in critical areas due to conflicting commands, ensuring rapid response and effective execution of police missions. In summary, the method of this application significantly improves the scheduling efficiency, safety, and coordination of police drone swarms in complex low-altitude airspace, effectively responding to emergency situations and overcoming many shortcomings of existing technologies. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a police drone swarm scheduling method provided in an embodiment of this application.

[0022] Figure 2 This is a schematic diagram of the structure of a police drone swarm dispatch system provided in an embodiment of this application.

[0023] Labeling Explanation: 210, Command Intent Parsing Module; 220, Airspace Information Receiving Module; 230, Collision Avoidance Logic Parameter Adjustment Module; 240, Cooperative Avoidance and Formation Adjustment Module. Detailed Implementation

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0025] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0026] With increasingly congested urban low-altitude airspace, police drone swarms face unprecedented challenges in mission execution. Traditional dispatching methods, when interacting with external civil low-altitude traffic management systems, suffer from protocol differences and data inconsistencies, leading to delays in airspace clearance acquisition and severely impacting the timeliness and flexibility of police missions. Furthermore, in emergency situations, conflicts may arise between manual intervention commands from the ground command center and automatic collision avoidance logic, causing instability in the swarm's flight status and even resulting in a "command deadlock," thus hindering effective response to emergency situations.

[0027] In this regard, firstly, referring to Figure 1 This application proposes a method for scheduling police drone swarms, including: S1. Obtain police mission instructions issued by the police drone dispatch system, perform intent analysis on the police mission instructions, and extract the mission intent and expected flight behavior; S2. Transform the mission intent and expected flight behavior into a semantic flight plan that can be understood by the external civil low-altitude traffic management system, submit the semantic flight plan to the external civil low-altitude traffic management system, and receive the airspace information returned by the external civil low-altitude traffic management system. S3. Obtain manual intervention instructions from the ground command center, identify the emergency intent of the manual intervention instructions, and comprehensively evaluate the manual intervention instructions based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and airspace information. Based on the comprehensive evaluation results, adjust the anti-collision logic parameters within the police drone cluster. S4. Obtain the behavioral change intentions of drones within the police drone cluster. Based on the behavioral change intentions, the current real-time situation of the police drone cluster, and the collision avoidance logic parameters, coordinate the drones within the police drone cluster to perform collaborative avoidance or formation adjustment.

[0028] This application aims to address the challenges faced by police drone swarms in complex low-altitude airspace, such as airspace permission delays, command conflicts, and difficulties in internal coordination, by introducing semantic flight plan conversion, comprehensive evaluation of manual intervention commands, and collaborative avoidance and formation adjustment mechanisms within the swarm. This will improve the mission response efficiency and flight safety of police drone swarms.

[0029] Police mission instructions refer to commands issued by the police drone dispatch system to instruct a swarm of police drones to perform specific police activities, such as patrolling, reconnaissance, and tracking. Mission intent refers to the core purpose and objective of the police mission instruction, such as emergency search and rescue or fixed-point surveillance. Expected flight behavior refers to the specific action plan, including flight path, altitude, speed, and time, that the drone swarm is expected to take to achieve the mission intent. Semantic flight planning involves converting the mission intent and expected flight behavior within the police system into a standardized, structured flight data format that can be understood and processed by an external civil low-altitude traffic management system, facilitating cross-system information exchange and airspace application. Airspace information refers to real-time data returned by the external civil low-altitude traffic management system regarding usage restrictions in specific airspaces, temporary avoidance zones, and the distribution of civil aircraft. Manual intervention instructions refer to commands issued directly to the police drone swarm by the ground command center in emergency situations to respond to unforeseen circumstances, which may contradict the automatic dispatch plan. Emergency intent refers to the urgency and purpose contained in the manual intervention instruction, such as immediate deployment or emergency avoidance. Real-time situational awareness refers to the dynamic information of all drones in a police drone swarm, including their current position, speed, attitude, battery level, mission status, and relative positions. Collision avoidance logic parameters refer to the algorithms and rules used within the police drone swarm to avoid collisions between drones or between drones and obstacles, such as safe distance thresholds and obstacle avoidance path calculation methods. Behavioral change intent refers to the intention of a drone in the swarm to change its original flight plan due to mission execution or human intervention, such as changing its heading, accelerating, decelerating, climbing, or descending. Cooperative avoidance refers to multiple drones in the swarm coordinating their flight paths to perform avoidance maneuvers when a potential collision risk is detected, ensuring the overall safety of the swarm. Formation adjustment refers to the reorganization of drone formations within the swarm based on mission requirements or changes in the external environment to optimize mission execution efficiency or improve flight stability.

[0030] The police drone swarm dispatching method proposed in this application first requires acquiring the police mission instructions issued by the police drone dispatching system and then parsing the instructions to extract the mission intent and expected flight behavior. For example, police mission instructions can be issued via text input, voice commands, or preset mission templates. Upon receiving the instructions, natural language processing technology can be used to parse the text or voice instructions, identifying keywords and phrases to determine the core mission intent, such as tracking a suspect vehicle or providing air support. Simultaneously, flight-related parameters are extracted from the instructions, such as the geographical coordinates of the target area, patrol route, flight altitude, and speed limits; these parameters constitute the expected flight behavior.

[0031] Subsequently, the aforementioned mission intent and anticipated flight behavior are converted into a semantic flight plan that can be understood by the external civil low-altitude traffic management system. This conversion process is crucial to ensuring that the police drone swarm can effectively interact with the broader low-altitude traffic management system. For example, priority information in the mission intent may be represented by numerical codes within the police system, but may need to be converted into textual descriptions such as emergency, priority, or routine in the civil system. Geographic coordinates in the anticipated flight behavior may need to be converted from the specific coordinate system within the police system to the internationally recognized WGS84 coordinate system and standardized to ensure consistency in accuracy. Timestamp information may also need to be synchronized with the external civil system via Network Time Protocol (NTP) or Precision Time Protocol (PTP) to eliminate time discrepancies. After the conversion is completed, the semantic flight plan is submitted to the external civil low-altitude traffic management system, and the system awaits the return of airspace information.

[0032] During missions conducted by a police drone swarm, the ground command center may issue manual intervention commands in response to emergencies. At this time, it is necessary to acquire these manual intervention commands and identify their emergency intent. For example, the manual intervention command may be sent via a dedicated communication link and include an emergency flag or a specific emergency code. Upon receiving the command, the system parses it to extract its emergency intent, such as immediately proceeding to a designated area or emergency evacuation. Then, based on the priority of the emergency intent, the real-time situation of the police drone swarm, and previously received airspace information, the manual intervention command is comprehensively evaluated. For example, the evaluation may include analyzing whether the urgency of the command matches the current emergency level, whether the execution of the command would pose a collision risk with other drones or civilian aircraft within the swarm, and whether the command would violate current no-fly zones. Based on the comprehensive evaluation results, the collision avoidance logic parameters within the police drone swarm are dynamically adjusted. For example, if the evaluation results indicate that the mission urgency is extremely high and the potential collision risk is controllable, the collision avoidance distance threshold can be appropriately relaxed to allow drones to fly more closely and thus reach the target area more quickly.

[0033] Finally, the system acquires the behavioral change intentions of drones within the police drone swarm and, based on these intentions, the real-time situation of the swarm, and adjusted collision avoidance logic parameters, coordinates drones within the swarm to perform cooperative avoidance or formation adjustments. For example, when a drone in the swarm needs to change course or accelerate due to an emergency mission, it generates a behavioral change intention and broadcasts it to the other drones in the swarm. Upon receiving this intention, the other drones, combining their own real-time position, speed, attitude, and the overall situational information of the swarm, and referring to the currently effective collision avoidance logic parameters, calculate their respective cooperative avoidance routes or formation reconfiguration instructions. This ensures that when a single drone changes behavior, the entire swarm can respond quickly and safely, avoiding collisions and maintaining overall mission efficiency.

[0034] In summary, this application effectively solves the problems faced by police drone swarms in complex low-altitude airspace, such as airspace permission delays, command conflicts, and difficulties in internal coordination, through semantic flight plan conversion, comprehensive evaluation of manual intervention commands, and collaborative avoidance and formation adjustment within the swarm. It provides an innovative solution for police drone swarms to perform missions efficiently and safely in increasingly busy urban low-altitude airspace.

[0035] Specifically, the steps of obtaining police mission instructions issued by the police drone dispatch system, performing intent analysis on the police mission instructions, and extracting mission intent and expected flight behavior can be further refined into the following methods.

[0036] The steps for obtaining police mission instructions from the police drone dispatch system, parsing the intent of the police mission instructions, and extracting the mission intent and expected flight behavior include: The predefined task semantic dictionary is used to match keywords in police task instructions to determine core intent labels and expected flight behavior parameters. The core intent labels are converted into task intents, and the expected flight behavior parameters are converted into expected flight behaviors. The predefined task semantic dictionary stores commonly used police task instructions and their corresponding core intent labels and expected flight behavior parameters.

[0037] Specifically, the predefined task semantic dictionary can be understood as a structured database or knowledge base, which pre-stores a large number of common police task instructions and their corresponding standardized core intent labels and expected flight behavior parameters. For example, for an emergency pursuit instruction, its core intent label might be defined as pursuit, while the expected flight behavior parameters might include high-speed flight, low-altitude tracking, and target lock. For an area patrol instruction, its core intent label might be defined as patrol, while the expected flight behavior parameters might include constant-speed flight, preset route, and periodic scanning. The dictionary is designed to cover various scenarios and instruction types that police drone swarms may encounter when performing tasks, ensuring the comprehensiveness and accuracy of the parsing.

[0038] In this context, matching keywords in police task instructions refers to the system using natural language processing technology to identify key information from the received police task instruction text that corresponds to entries in a predefined task semantic dictionary. For example, lexical analysis, syntactic analysis, or rule-based pattern matching algorithms can be used to identify key elements such as verbs, nouns, time, and location in the instructions. Once keywords are identified, the system compares them with preset keywords in the dictionary to determine the core intent label and expected flight behavior parameters that best match the current instruction.

[0039] Furthermore, the core intent label is an abstract summary of the police mission instructions, representing the fundamental purpose or nature of the mission, such as reconnaissance, search and rescue, surveillance, or strike. The expected flight behavior parameters detail the specific flight modes and operational requirements that the drone swarm should adopt to achieve this core intent, such as flight speed, altitude, flight path type, formation, and sensor usage patterns. By converting the matched core intent label into mission intent and the expected flight behavior parameters into expected flight behavior, the transformation from unstructured police mission instructions to structured, executable flight plan elements is achieved.

[0040] Specifically, the steps described above for translating mission intent and anticipated flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system include: The task priority in the mission intent is mapped to a text description accepted by the external civil low-altitude traffic management system through the task priority encoding inside the police system; The geographic coordinates in the expected flight behavior are standardized based on the internationally accepted coordinate system. The timestamps in the expected flight behavior are synchronized based on the network time protocol or the precision time protocol in order to synchronize with the external civil low-altitude traffic management system. The mapped text description, standardized geographic coordinates, and synchronized timestamps are integrated and converted into a semantic flight plan.

[0041] The process of mapping task priorities from the mission intent to a text description acceptable to the external civil low-altitude traffic management system (CALS) involves addressing potential differences in task priority definitions and expressions between the police drone dispatch system and the CALS. This conversion ensures the external system can correctly understand and process the urgency of the police mission. Specifically, the task priority codes within the police system, such as "urgent," "high," "medium," and "low," are mapped to standardized text descriptions acceptable to the CALS, such as "emergency flight," "priority passage," or "regular flight." This aims to eliminate semantic differences between systems and ensure accurate transmission of task priorities in cross-system communication.

[0042] Furthermore, standardizing the geographic coordinates in the anticipated flight behavior based on an internationally accepted coordinate system can be understood as converting the geographic coordinate format used internally by the police drone dispatch system into an internationally widely accepted coordinate system to ensure the interoperability and accuracy of geographic location information between different systems. For example, the police system may use a local coordinate system or a specific projected coordinate system, while external civilian low-altitude traffic management systems typically use internationally accepted geodetic coordinate systems such as WGS84. Through standardization, positioning errors caused by coordinate system mismatch can be eliminated, ensuring the accurate sharing of flight path and airspace information.

[0043] Furthermore, synchronizing timestamps in anticipated flight activities based on Network Time Protocol (NTP) or Precision Time Protocol (PTP) to synchronize with external civil low-altitude traffic management systems (CALS) ensures that the flight plans of police drone swarms are consistent with the time reference of CALS, avoiding safety hazards or scheduling conflicts caused by time discrepancies. Specifically, using time synchronization protocols such as NTP or PTP ensures that timestamps such as takeoff time, arrival time, and segment time in the flight plans of police drone swarms are strictly synchronized with the time of external CALS, aiming to provide a unified time reference for accurate airspace management and conflict detection.

[0044] Therefore, integrating the mapped text description, standardized geographic coordinates, and synchronized timestamps into a semantic flight plan involves encapsulating the key information, such as the task priority text description, geographic coordinates, and timestamps, after the aforementioned transformations and standardization processes, according to the specific data format and semantic structure required by external civil low-altitude traffic management systems. For example, it can be integrated into a flight plan message format conforming to the regulations of the International Civil Aviation Organization (ICAO) or relevant national aviation management agencies, such as JSON or XML structured data. The aim is to create a complete, accurate, and semantically clear flight plan that external systems can directly parse, understand, and use for airspace management and flight scheduling.

[0045] In some embodiments described above, receiving airspace information returned by an external civil low-altitude traffic management system is proposed. However, during implementation, if the received airspace information is not further parsed and transformed, the police drone swarm may not be able to effectively integrate it into its collision avoidance system, thereby affecting the safety and efficiency of mission execution. Specifically, the airspace information provided by the external civil low-altitude traffic management system is usually general and may not be directly applicable to the specific collision avoidance logic and real-time decision-making needs of the police drone swarm. Without proper transformation, the police drone swarm will find it difficult to quickly and accurately adjust its flight strategy when facing complex airspace restrictions and emergencies, potentially increasing the risk of collision or reducing mission response speed.

[0046] In response, this application further proposes a method for scheduling police drone swarms, which includes a step performed after receiving the airspace information returned by the aforementioned external civil low-altitude traffic management system: Analyze airspace information to obtain no-fly altitudes, temporary avoidance zones, and areas with high concentrations of civilian aircraft; No-fly altitudes, temporary avoidance zones, and densely populated areas of civilian aircraft are converted into dynamic airspace boundary parameters or obstacle avoidance behavior adjustment instructions that can be identified by the police drone collision avoidance system.

[0047] Specifically, parsing airspace information refers to the structured processing and content extraction of raw airspace data received from external civil low-altitude traffic management systems. This airspace information can contain data in various formats and types, such as XML, JSON, or specific binary protocols. Through parsing, information crucial to the flight safety of police drone swarms can be accurately identified and extracted, including no-fly altitudes, temporary avoidance zones, and densely populated areas of civil aircraft. No-fly altitudes refer to the minimum or maximum altitude restrictions within a specific airspace that drones are prohibited from flying over; temporary avoidance zones are airspace areas temporarily designated due to specific events (such as temporary air traffic control, emergency rescue operations, etc.) that police drones must avoid; and densely populated areas of civil aircraft are areas with frequent civil aircraft activity, where police drones need to pay special attention to collision avoidance.

[0048] Furthermore, converting the aforementioned no-fly altitudes, temporary avoidance zones, and densely populated areas of civilian aircraft into dynamic airspace boundary parameters or obstacle avoidance behavior adjustment commands that can be recognized by the police drone collision avoidance system means transforming these extracted general airspace restriction information into commands or parameters that the police drone swarm's internal collision avoidance system can directly understand and execute. Dynamic airspace boundary parameters may include three-dimensional coordinate points, altitude limits, time windows, etc., used to update the available flight space model of the drone swarm in real time. Obstacle avoidance behavior adjustment commands can be specific flight path correction suggestions, speed limits, altitude adjustment commands, or triggering conditions for specific avoidance actions. For example, when a densely populated area of ​​civilian aircraft is identified, commands can be generated requiring the swarm drones to reduce their flight altitude or change their course to avoid the area.

[0049] In some preferred embodiments, a specific example is given below. Suppose a police drone dispatch system receives an emergency mission instruction requiring a swarm of police drones to conduct reconnaissance in a certain area. After submitting a semantic flight plan to an external civil low-altitude traffic management system and receiving airspace information, the method immediately initiates a parsing step. For example, the received airspace information may include a temporary no-fly zone with coordinates ranging from [X1,Y1] to [X2,Y2], an altitude restriction of 0 meters to 150 meters, and an area with frequent civilian helicopter activity near the no-fly zone.

[0050] At this point, the system analyzes this information, identifying a no-fly altitude below 150 meters, a temporary avoidance zone of [X1,Y1]-[X2,Y2], and marking the area where civilian helicopters operate as a dense civilian aircraft zone. This information is then converted into dynamic airspace boundary parameters for the police drone collision avoidance system. For example, the 150-meter no-fly altitude is converted into an adjustment parameter for the upper limit of the swarm drone's flight altitude; the temporary avoidance zone is defined as a high-priority avoidance zone in the swarm path planning algorithm; and the dense civilian aircraft zone may trigger obstacle avoidance behavior adjustment commands, such as requiring swarm drones to automatically reduce speed, increase drone spacing, and activate additional sensors for target identification and tracking when flying near that area to ensure safety. In this way, the police drone swarm can integrate external airspace restrictions into its flight decisions in real time when performing missions, thereby effectively avoiding risks.

[0051] In some of the embodiments described above in this application, a scheme is proposed to comprehensively evaluate manual intervention commands based on the priority of the emergency intent, the real-time situation of the current police drone swarm, and airspace information, and to adjust the collision avoidance logic parameters based on the evaluation results. However, in practical applications, this comprehensive evaluation process may lack quantitative consideration of the urgency of the task and potential safety risks, resulting in insufficient accuracy and operability of the evaluation results. Especially in complex and ever-changing police situation environments, it is difficult to effectively guide the fine-tuning of collision avoidance logic parameters. If the above problems are not solved, police drone swarms may face higher safety risks or reduced task efficiency when performing emergency tasks due to inaccurate evaluations. To address this, this application further proposes a more refined comprehensive evaluation method, which improves the accuracy of the evaluation and the scientific nature of the decision-making by introducing the current police situation level and quantifying the urgency of the task and the probability of potential collisions.

[0052] Based on the above method, the steps for comprehensively evaluating manual intervention commands according to the priority of the emergency intent, the real-time situation of the current police drone swarm, and airspace information include: The system obtains the current alarm level and, based on the current alarm level, the priority of the emergency intent, the real-time situation of the current police drone swarm, and the dynamic airspace boundary parameters or obstacle avoidance behavior adjustment instructions, comprehensively assesses the urgency and potential safety risks of the manual intervention instruction. The system calculates the task urgency score and the potential collision probability as the comprehensive assessment result.

[0053] Specifically, obtaining the current alert level refers to the alert level that the system acquires in real time or that is input by the ground command center. For example, it can be categorized as a general alert, a major alert, or a critical alert, etc., with the aim of providing important weighting factors for subsequent mission urgency assessment. The mission urgency score can be understood as a numerical assessment of the urgency of the mission represented by the manual intervention command. This score comprehensively considers the priority of the emergency intent, the current alert level, and the mission's impact on the real-time situation of the police drone swarm. In practical applications, the mission urgency score can be calculated using a preset scoring model or expert system. For example, a weighted average method can be used, assigning different weights to factors such as alert level and emergency intent priority for cumulative calculation. Furthermore, the potential collision probability refers to the likelihood that the police drone swarm will collide with other drones, civilian aircraft, or airspace obstacles within the swarm after executing the manual intervention command. Its purpose is to quantitatively assess the safety risks during mission execution. The calculation of potential collision probability can be based on real-time situational data of the current police drone swarm (such as position, speed, and heading), dynamic airspace boundary parameters, or obstacle avoidance behavior adjustment commands, through methods such as predictive trajectory analysis and collision risk models. For example, Monte Carlo simulation or methods based on probability density functions can be used to predict collision risks over a future period.

[0054] In practical applications, there is still room for optimization in how to accurately trigger adjustments to collision avoidance logic parameters based on complex comprehensive assessment results, especially mission urgency scores and potential collision probabilities, while ensuring that the adjustments meet both urgent mission requirements and effectively control flight safety risks. Without clearly defined triggering conditions and adjustment strategies, the collision avoidance system may adjust at inappropriate times or by inappropriate magnitudes, thereby affecting the effective execution of police missions or introducing unnecessary safety hazards.

[0055] In this regard, this application further proposes the following steps for adjusting the anti-collision logic parameters within the police drone swarm based on the comprehensive evaluation results: When the mission urgency score is higher than the preset urgency threshold and the potential collision probability is lower than the preset collision threshold, the collision avoidance distance threshold within the police drone cluster is adjusted, a temporary safe flight path is generated, and real-time risk alerts are provided to the ground command center.

[0056] Specifically, the aforementioned "task urgency score higher than the preset urgency threshold and potential collision probability lower than the preset collision threshold" means that after comprehensively evaluating a manual intervention command, the system obtains a quantified task urgency score and a potential collision probability. The task urgency score reflects the urgency of the manual intervention command, while the preset urgency threshold is a configurable parameter used to define the urgency level of the task, which can be dynamically set based on factors such as the level of the incident and the task type. The potential collision probability represents the likelihood of a collision occurring under the current real-time situation and airspace information of the police drone swarm; the preset collision threshold is an acceptable upper limit for collision risk. Only when the urgency of the task reaches a certain level (i.e., the task urgency score is higher than the preset urgency threshold) and the collision risk is within a controllable range (i.e., the potential collision probability is lower than the preset collision threshold) will the system trigger adjustments to the anti-collision logic parameters to avoid unnecessary parameter changes in extremely high-risk or non-emergency situations.

[0057] Adjusting the collision avoidance distance threshold within a police drone swarm can be understood as the system dynamically modifying the minimum safe distance between drones or between a drone and an obstacle, based on the aforementioned judgment conditions. For example, when performing highly urgent and manageable-risk tasks, the collision avoidance distance threshold can be appropriately reduced, allowing drones to fly in tighter formations, thereby improving mission response speed and efficiency. Conversely, in higher-risk or non-emergency situations, the threshold can be maintained or increased to ensure a higher safety margin.

[0058] In practical applications, generating a temporary safe flight path refers to the system, while adjusting the collision avoidance distance threshold, replanning a temporary, safe flight path for the affected drones based on new collision avoidance parameters, the real-time situation of the current police drone swarm, and airspace information. This flight path aims to guide the drones to effectively avoid potential collision risks while meeting mission requirements and adapting to the new collision avoidance logic parameters. Furthermore, providing real-time risk alerts to the ground command center is a crucial safety mechanism. Even if parameters are adjusted and a safe flight path is generated in an emergency, the system will still report the current operational status, parameter adjustments, and potential risks to the ground command center, enabling human operators to supervise and intervene when necessary, thereby achieving human-machine collaborative safety management.

[0059] In some of the embodiments described above in this application, although a scheme is proposed to coordinate drones in a police drone swarm to perform collaborative avoidance or formation adjustment based on behavioral change intentions, real-time situation and collision avoidance logic parameters, in practical applications, when a drone in the swarm needs to urgently adjust its flight behavior due to a sudden situation or manual command, the lack of a clear internal communication and coordination mechanism may cause other drones in the swarm to be unable to respond in a timely and accurate manner, thereby affecting the overall flight safety and mission execution efficiency of the swarm.

[0060] In response, this application further proposes the steps for obtaining the behavioral change intentions of drones within a police drone swarm, and coordinating drones within the swarm to perform cooperative avoidance or formation adjustments based on the behavioral change intentions, the current real-time situation of the police drone swarm, and the collision avoidance logic parameters. When a drone in a police drone swarm adjusts its flight behavior due to performing an emergency mission or due to human intervention, it acquires the intention to change its behavior and broadcasts the intention to change its behavior to other drones in the police drone swarm. Based on the behavioral change intent, the real-time situation of the current police drone swarm, and the collision avoidance logic parameters, the system calculates collaborative avoidance flight paths or formation reconfiguration instructions for neighboring drones.

[0061] Specifically, when a drone within a police drone swarm needs to immediately change its course, speed, or altitude to perform a new mission, for example, due to receiving an emergency instruction from the ground command center, or needing to perform an emergency maneuver to avoid a sudden obstacle, its flight behavior will be adjusted. In this case, the drone will proactively acquire its own intention to change its behavior. This intention can be understood as specific information about the drone's planned flight behavior adjustments, such as the new target location, estimated arrival time, new speed, new altitude, and the reason for the adjustment. Upon acquiring this intention, the drone will immediately broadcast it to the other drones in the swarm. The purpose of this broadcast is to ensure that all relevant drones in the swarm, especially its neighboring drones, can promptly perceive this important change in flight status, thereby providing the necessary information basis for subsequent collaborative decision-making.

[0062] Upon receiving a broadcast intent to change behavior, neighboring drones within the cluster will perform real-time calculations and decisions based on this intent, the current real-time situation of the police drone cluster, and pre-defined collision avoidance logic parameters. The current real-time situation of the police drone cluster includes the real-time position, speed, attitude, and mission status of all drones within the cluster. Collision avoidance logic parameters define rules for maintaining safe distances and avoidance strategies among drones within the cluster. By comprehensively analyzing this information, neighboring drones can calculate the optimal cooperative avoidance flight path or formation reconfiguration command. A cooperative avoidance flight path refers to a new flight path planned to avoid collisions with drones adjusting their behavior; a formation reconfiguration command refers to adjustments made to their own position, speed, etc., to maintain the overall formation structure and mission objectives of the cluster. The aim is to efficiently respond to changes in the behavior of individual drones while ensuring the overall safety of the cluster, maintaining cluster coordination and mission continuity.

[0063] In some of the embodiments described above in this application, a scheme is proposed to calculate cooperative avoidance routes or formation reconfiguration commands for neighboring UAVs based on behavioral change intentions, real-time situation, and collision avoidance logic parameters. However, in its implementation, how to efficiently and accurately identify neighboring UAVs that need cooperative avoidance or formation adjustment is crucial to ensuring the real-time performance and security of cluster scheduling. If the identification efficiency of neighboring UAVs is low or the accuracy is insufficient, it may lead to delayed response of cooperative avoidance or formation adjustment, or even trigger potential collision risks.

[0064] In response, this application further proposes the following steps for identifying nearby drones: Maintain a real-time cluster member location list. Based on the real-time cluster member location list, identify all neighboring drones within a preset range of the drone with behavioral change intentions. All police drones report their real-time location, speed, attitude, battery level, and mission status at a preset frequency. The real-time cluster member location list is updated in real time based on the real-time location, speed, attitude, battery level, and mission status.

[0065] Specifically, the real-time cluster member location list can be understood as a dynamic database or data structure that stores key operational data for all drones within the police drone cluster. This list is maintained by all police drones in the cluster reporting their real-time location, speed, attitude, battery level, and mission status to the cluster scheduling system or distributed cluster management node at a preset frequency, such as several times or dozens of times per second. Real-time location can include geographic coordinates such as latitude, longitude, and altitude; speed can include horizontal and vertical speeds; attitude can include roll, pitch, and yaw angles; battery level is used to assess the drone's endurance; and mission status indicates the type or stage of the mission the drone is currently performing. The real-time cluster member location list is updated in real time based on these reports, ensuring that the information stored always reflects the latest situation of the cluster.

[0066] The identification of all neighboring drones within a preset range of a drone exhibiting behavioral change intent refers to the system defining a preset geographical range or safety distance centered on a drone in the cluster when the drone demonstrates behavioral change intent due to an emergency mission or human intervention. Then, by querying the real-time cluster member location list, all other drones within this preset range are selected; these selected drones are defined as neighboring drones. The preset range can be dynamically adjusted based on factors such as mission type, airspace environment, drone performance, and collision avoidance logic parameters. For example, in high-speed flight or complex airspace, the preset range can be appropriately expanded to provide more reaction time.

[0067] Specifically, the constituent elements of the aforementioned behavioral changes are crucial for the coordinated evasion or formation adjustment of police drone swarms.

[0068] The intent to change behavior includes the drone's unique identifier, new target location, estimated time of arrival, and new speed and altitude.

[0069] Specifically, a drone's unique identifier refers to a code or name used to uniquely identify a specific drone within a police drone swarm. Its purpose is to ensure that all drones within the swarm can accurately identify the source drone issuing the behavioral change intention. The new target location refers to the geographic coordinates, such as longitude, latitude, and altitude, that the drone plans to reach after adjusting its flight behavior. Its purpose is to clearly define the drone's new spatial target. The estimated time of arrival refers to the time the drone is expected to reach the new target location. Its purpose is to provide a time reference for other drones within the swarm to facilitate synchronized coordinated avoidance or formation adjustments. The new speed refers to the planned flight speed of the drone after adjusting its flight behavior. Its purpose is to reflect the dynamic characteristics of the drone when performing new tasks or avoiding obstacles. The new altitude refers to the planned flight altitude of the drone after adjusting its flight behavior. Its purpose is to ensure safe vertical separation and airspace compliance for the drones.

[0070] While the aforementioned method for scheduling police drone swarms can effectively schedule police drone swarms, its execution efficiency and reliability may be affected if a specific system architecture is lacking, and modular management and maintenance will be difficult to achieve.

[0071] Regarding this, secondly, refer to Figure 2 This application proposes a police drone swarm scheduling system for executing the aforementioned police drone swarm scheduling method. The system includes: The instruction intent parsing module 210 is used to obtain police task instructions issued by the police drone dispatch system, perform intent parsing on the police task instructions, and extract the task intent and expected flight behavior. The airspace information receiving module 220 is used to convert mission intent and expected flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system, submit the semantic flight plan to the external civil low-altitude traffic management system, and receive airspace information returned by the external civil low-altitude traffic management system. The anti-collision logic parameter adjustment module 230 is used to obtain manual intervention instructions from the ground command center, identify the emergency intent of the manual intervention instructions, comprehensively evaluate the manual intervention instructions based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and airspace information, and adjust the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results. The collaborative avoidance and formation adjustment module 240 is used to obtain the behavioral change intentions of drones in the police drone cluster, and coordinate the drones in the police drone cluster to perform collaborative avoidance or formation adjustment based on the behavioral change intentions, the current real-time situation of the police drone cluster and the collision avoidance logic parameters.

[0072] Specifically, the instruction intent parsing module 210 is configured to acquire police mission instructions issued by the police drone dispatch system and perform intent parsing on these instructions to extract the mission intent and expected flight behavior. This module can use natural language processing technology or keyword matching based on a predefined task semantic dictionary to transform unstructured mission instructions into structured data that the system can process, ensuring accurate understanding of the mission information.

[0073] The airspace information receiving module 220 is configured to convert mission intent and anticipated flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system, submit the semantic flight plan to the external civil low-altitude traffic management system, and receive airspace information returned by the external civil low-altitude traffic management system. This module is responsible for data interaction and protocol conversion between the police system and the external civil air traffic control system, ensuring that the flight activities of the police drone swarm can be legally and safely monitored by the external air traffic control system.

[0074] The anti-collision logic parameter adjustment module 230 is configured to acquire manual intervention commands from the ground command center, identify the emergency intent of the manual intervention commands, comprehensively evaluate the manual intervention commands based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and airspace information, and adjust the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results. This module is crucial for the safe operation of the system; it can dynamically respond to emergencies and manual interventions, balancing mission execution efficiency and flight safety by adjusting the anti-collision parameters.

[0075] The collaborative avoidance and formation adjustment module 240 is configured to acquire the behavioral change intentions of drones within the police drone cluster. Based on the behavioral change intentions, the current real-time situation of the police drone cluster, and the collision avoidance logic parameters, it coordinates the drones within the cluster to perform collaborative avoidance or formation adjustment. This module ensures that drones within the cluster can efficiently and safely perform collaborative actions when facing emergencies or mission adjustments, avoiding internal collisions and maintaining the overall mission capability of the cluster.

[0076] The solution proposed in this application effectively addresses the efficiency and reliability issues that may arise during actual deployment and operation of the aforementioned police drone swarm scheduling method by decoupling its functions into independent system modules. The instruction intent parsing module 210, serving as the system's entry point, is responsible for converting raw task instructions into system-recognizable intents and behaviors, laying the foundation for subsequent flight plan generation. The airspace information receiving module 220 acts as a bridge for interaction with external air traffic control systems, ensuring that the flight activities of the police drone swarm comply with external airspace management regulations and acquiring necessary airspace safety information. The anti-collision logic parameter adjustment module 230, upon receiving manual intervention instructions, can intelligently evaluate based on multi-source information and dynamically adjust the swarm's anti-collision strategy to respond to emergencies. Finally, the collaborative avoidance and formation adjustment module 240 is responsible for the refined management within the swarm, ensuring that the entire swarm can quickly and safely coordinate its response when individual drone behaviors change. It is precisely this modular design that enables the complex scheduling method to be executed efficiently and stably.

[0077] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A police unmanned aerial vehicle cluster scheduling method, characterized in that, include: The system acquires police mission instructions issued by a police drone dispatch system, performs intent parsing on the instructions, and extracts the mission intent and expected flight behavior. The mission intent refers to the core purpose and objective of the police mission instruction, and the expected flight behavior refers to the specific action plan, including flight path, altitude, speed, and time, that the drone swarm is expected to take to achieve the mission intent. Specifically, a predefined task semantic dictionary is used to match keywords in the police mission instructions to determine core intent tags and expected flight behavior parameters. The core intent tags are then converted into the mission intent, and the expected flight behavior parameters are converted into the expected flight behavior. The mission intent and expected flight behavior are converted into a semantic flight plan that can be understood by an external civil low-altitude traffic management system. The semantic flight plan is submitted to the external civil low-altitude traffic management system, and airspace information returned by the external civil low-altitude traffic management system is received. The system obtains manual intervention instructions from the ground command center, identifies the emergency intent of the manual intervention instructions (the emergency intent refers to the urgency and purpose contained in the manual intervention instructions), comprehensively evaluates the manual intervention instructions based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and the airspace information, and adjusts the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results. The system obtains the behavioral change intentions of drones within a police drone cluster. Based on the behavioral change intentions, the real-time situation of the current police drone cluster, and the anti-collision logic parameters, it coordinates the drones within the police drone cluster to perform cooperative avoidance or formation adjustment. The real-time situation of the current police drone cluster includes the real-time position, speed, attitude, and mission status of all drones within the cluster. The method further includes steps performed after receiving airspace information returned by the external civil low-altitude traffic management system: Analyze the airspace information to obtain no-fly altitudes, temporary avoidance zones, and areas with high concentrations of civilian aircraft; The no-fly altitude, temporary avoidance zone, and densely populated area of ​​civilian aircraft are converted into dynamic airspace boundary parameters or obstacle avoidance behavior adjustment instructions that can be identified by the police drone collision avoidance system. The step of comprehensively evaluating the manual intervention command based on the priority of the emergency intent, the real-time situation of the current police drone swarm, and the airspace information includes: The current alarm level is obtained. Based on the current alarm level, the priority of the emergency intention, the real-time situation of the current police drone cluster, and the dynamic airspace boundary parameters or the obstacle avoidance behavior adjustment instructions, the task urgency and potential safety risks of the manual intervention instructions are comprehensively evaluated, and the task urgency score and potential collision probability are calculated as the comprehensive evaluation result. Among them, obtaining the current alarm level refers to the alarm level that the system obtains in real time or that is input by the ground command center. The mission urgency score is a numerical value that quantifies the urgency of the mission represented by the manual intervention command. The potential collision probability refers to the possibility that the police drone swarm will collide with other drones, civilian aircraft or airspace obstacles within the swarm after executing the manual intervention command. The steps for adjusting the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results include: When the urgency score of the mission is higher than the preset urgency threshold and the potential collision probability is lower than the preset collision threshold, the collision avoidance distance threshold within the police drone cluster is adjusted, a temporary safe flight path is generated, and real-time risk alerts are provided to the ground command center. 2.The police unmanned aerial vehicle cluster scheduling method according to claim 1, characterized in that, The predefined task semantic dictionary stores commonly used police task instructions and their corresponding core intent tags and expected flight behavior parameters. 3.The police unmanned aerial vehicle cluster scheduling method according to claim 1, characterized in that, The step of converting the mission intent and expected flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system includes: The task priority in the stated task intent is mapped to a text description accepted by the external civil low-altitude traffic management system through task priority encoding within the police system. The geographic coordinates in the expected flight behavior are standardized based on an internationally accepted coordinate system. The timestamps in the expected flight behavior are synchronized based on the Network Time Protocol or Precision Time Protocol in order to synchronize with the external civil low-altitude traffic management system. The mapped text description, the standardized geographic coordinates, and the synchronized timestamp are integrated and converted into the semantic flight plan.

4. The police unmanned aerial vehicle cluster scheduling method according to claim 1, characterized in that, The steps of obtaining the behavioral change intentions of drones within the police drone cluster, and coordinating drones within the cluster to perform cooperative avoidance or formation adjustment based on the behavioral change intentions, the real-time situation of the current police drone cluster, and the collision avoidance logic parameters, include: When a drone in a police drone cluster adjusts its flight behavior due to performing an emergency mission or due to human intervention, the intention of the behavior change is obtained and broadcast to other drones in the police drone cluster. Based on the behavioral change intent, the real-time situation of the current police drone cluster, and the collision avoidance logic parameters, cooperative avoidance flight path or formation reconstruction instructions are calculated for neighboring drones.

5. The police unmanned aerial vehicle cluster scheduling method according to claim 4, characterized in that, The nearby drone is identified through the following steps: Maintain a real-time cluster member location list, and identify all neighboring drones within a preset range of the drone with the intention of the behavioral change based on the real-time cluster member location list. All police drones report their real-time location, speed, attitude, battery level, and mission status at a preset frequency. The real-time cluster member location list is updated in real time according to the real-time location, speed, attitude, battery level, and mission status.

6. The police unmanned aerial vehicle cluster scheduling method according to claim 4, characterized in that, The behavioral change intentions include the drone's unique identifier, new target location, estimated time of arrival, and new speed and altitude.

7. A police drone swarm dispatching system for performing the police drone swarm dispatching method according to any one of claims 1 to 6, characterized in that, The system includes: The instruction intent parsing module is used to obtain police task instructions issued by the police drone dispatch system, perform intent parsing on the police task instructions, and extract the task intent and expected flight behavior. The airspace information receiving module is used to convert the mission intent and expected flight behavior into a semantic flight plan that can be understood by an external civil low-altitude traffic management system, submit the semantic flight plan to the external civil low-altitude traffic management system, and receive airspace information returned by the external civil low-altitude traffic management system. The anti-collision logic parameter adjustment module is used to obtain manual intervention instructions from the ground command center, identify the emergency intent of the manual intervention instructions, comprehensively evaluate the manual intervention instructions based on the priority of the emergency intent, the real-time situation of the current police drone cluster, and the airspace information, and adjust the anti-collision logic parameters within the police drone cluster based on the comprehensive evaluation results. The collaborative avoidance and formation adjustment module is used to obtain the behavioral change intentions of drones within the police drone cluster, and coordinate the drones within the police drone cluster to perform collaborative avoidance or formation adjustment based on the behavioral change intentions, the real-time situation of the current police drone cluster, and the anti-collision logic parameters.

Images