Conflict warning method and system for manned-unmanned aircraft fusion operation

Abstract

The application relates to a manned-unmanned aircraft fusion operation conflict warning method and system, which comprises the following steps: acquiring multi-source monitoring data of a local aircraft and an invading aircraft in each sampling period, filtering and fusing to obtain a fusion state vector and a covariance matrix, extracting a performance error component, synthesizing an error safety margin, constructing a basic clearance protection zone with the invading aircraft fusion position as the center, dynamically expanding the protection zone by using the error safety margin to obtain a dynamic clearance protection zone, constructing a relative collision cone with a dynamic horizontal radius as a safety interval to perform first-stage pre-screening, entering second-stage fine detection, calculating in-out protection zone time and determining whether there is a clearance conflict, comparing a conflict entering time with an adaptive warning threshold value, solving a common tangent line of a local aircraft maneuvering track and a dynamic protection zone protection circle when near-term warning occurs, directly obtaining a heading and ground speed conflict boundary value, determining a conflict state of each detection interval, and generating a visual conflict prevention zone.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to a conflict alarm method and system for manned-unmanned aerial vehicle (UAV) integrated operation. Background Technology

[0002] Because drones lack the real-time situational awareness and decision-making capabilities of airborne pilots, and because their communication links are subject to delays and their navigation and positioning accuracy is limited by sensor performance, the risk of flight conflicts with manned aircraft in the merged airspace increases significantly.

[0003] Currently, in research on drone conflict detection and alarm, a representative technical solution is the fixed threshold conflict detection and alarm method based on the closest point distance. This method acquires the position and velocity data of the local drone and the intruding drone at each sampling moment. Assuming both drones maintain their current speed and direction while moving in a uniform straight line, it calculates the predicted distance between the two drones when they reach their horizontal closest point within a predicted time period based on their relative position and velocity vectors. This predicted distance is compared with a preset fixed safety interval threshold. If the predicted distance is less than the fixed safety interval threshold, an alarm is triggered; if the predicted distance is greater than or equal to the fixed safety interval threshold, the situation is considered safe. Some improved solutions introduce a preset fixed alarm time threshold, triggering an alarm even when the estimated time for the two drones to reach their closest point is less than the fixed alarm time threshold.

[0004] The aforementioned technologies employ fixed safety interval thresholds and fixed alarm time thresholds, failing to consider the differences and real-time changes in sensor measurement accuracy under different flight scenarios. When airborne sensors experience increased positioning errors due to environmental interference, the distance to the closest point calculated based on the measured value may deviate significantly from the true value. Fixed thresholds cannot adapt to such error fluctuations, leading to an increase in the missed alarm rate or an excessively high false alarm rate, resulting in low alarm reliability.

[0005] Based on this, this application provides a conflict alarm method and system for manned-unmanned aerial vehicle (UAV) integrated operation. Summary of the Invention

[0006] To address the issue that using fixed safety interval thresholds and fixed alarm time thresholds fails to consider the differences and real-time changes in sensor measurement accuracy under different flight scenarios, and when the positioning error of airborne sensors increases due to environmental interference, the distance to the nearest point calculated based on the measurement value may deviate significantly from the true value. Fixed thresholds cannot adapt to such error fluctuations, leading to an increase in the false alarm rate or an excessively high false alarm rate, resulting in low alarm reliability, this application provides a conflict alarm method and system for manned and unmanned aerial vehicle (UAV) integrated operation.

[0007] Firstly, this application provides a conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation, which adopts the following technical solution: including: Based on the sampling period, multi-source data from the local machine and the intrusion machine are acquired and fused through Kalman filtering to obtain the fused state vector and covariance matrix. Based on the covariance matrix, performance error components of the manned and unmanned ends are constructed, and each performance error component is combined into a core error loop and a limit error loop, which together constitute the error safety margin. Based on the fusion state vector, a basic airspace protection zone is constructed with the fusion position of the intruder as the center; the horizontal radius and vertical half height of the basic protection zone are dynamically expanded using the error safety margin to obtain a dynamic airspace protection zone. Using the dynamic horizontal radius of the dynamic airspace protection zone as a safety interval, a relative collision cone region is constructed; if the relative velocity vector of the machine is not in the relative collision cone region, it is considered safe; otherwise, the time for the machine to enter and exit the dynamic airspace protection zone is calculated to determine whether there is an airspace conflict and obtain the conflict entry time. Based on the conflict entry time and a preset threshold range, the alarm level is determined, including: recent alarm level, mid-term warning level, and security level; When the alarm level is determined to be recent, the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone are used to calculate the conflict boundary value of the heading and ground speed. Combined with the vertical boundary conditions, the conflict of the divided detection zones is determined. Zones with conflicts are assigned to the warning zone, and zones without conflicts are assigned to the safe maneuver zone.

[0008] Preferably, the multi-source data includes: ADS-B broadcast data, satellite navigation data, airborne sensor detection data, C2 link data, and meteorological observation data; The process involves constructing performance error components for both manned and unmanned aircraft based on the covariance matrix, and then synthesizing these performance error components into a core error loop and a limit error loop, which together constitute the error safety margin, including: Using the Kalman filter, position and velocity information from ADS-B broadcast data are used as state prediction input, and position and velocity information from satellite navigation data and airborne sensor detection data are used as observation update input for recursive fusion to obtain the fused state vector and the covariance matrix. The fused state vector includes the local fused position, local fused velocity, intruder fused position, and intruder fused velocity. Based on the submatrix corresponding to the local machine in the covariance matrix, the position variance and velocity variance are extracted. Combined with the wind speed variance in the meteorological observation data, the UAV end performance error components are constructed. The UAV end performance error components include satellite navigation positioning error and wind speed yaw error. Based on the submatrix corresponding to the intruder in the covariance matrix, the position variance and velocity variance are extracted. Combined with the preset communication performance accuracy parameters, navigation performance accuracy parameters and monitoring performance refresh rate parameters, the manned aircraft performance error component is constructed. The manned aircraft performance error component includes communication performance positioning error, navigation performance positioning error and monitoring performance positioning error. Based on the error values ​​of the UAV-side performance error components and the manned aircraft-side performance error components, the error boundary value of the core error loop is obtained; based on the error boundary value of the core error loop, the maximum value of each error component, and a preset amplification factor, the error boundary value of the limit error loop is obtained.

[0009] Preferably, the basic airspace protection zone is constructed based on the fusion state vector, with the fusion position of the intruder as the center; the horizontal radius and vertical half-height of the basic airspace protection zone are dynamically expanded using the error safety margin to obtain a dynamic airspace protection zone, including: A cylindrical protection zone is constructed with the intruder's fusion location as the center, a preset horizontal minimum separation threshold as the radius, and a preset vertical minimum separation threshold as the half-height. Based on a preset time threshold, the cylindrical protection zone is horizontally extended forward along the relative motion direction of the two aircraft to form the basic airspace protection zone. The position error boundary and velocity error boundary are extracted from the error safety margin; the dynamic time variable is calculated, which is the smaller value between the preset prediction time window and the corrected closest point time. The corrected closest point time is the estimated time to reach the nearest boundary of the protected area after correction based on the fused state vector and the error safety margin. The dynamic horizontal radius is obtained by superimposing the horizontal radius with the position error boundary and the sum of the dynamic time variable multiplied by the velocity error boundary; the dynamic vertical half-height is obtained by superimposing the vertical position error boundary with the dynamic time variable multiplied by the vertical velocity error boundary, thus forming the dynamic airspace protection zone. Specifically, the dot product of the relative velocity vector and the relative position vector of the two machines is calculated. When the dot product is not less than zero, the two machines are in a state of distance. When the dot product is less than zero and the relative distance between the two machines is greater than a preset warning distance threshold, the two machines are in a state of slow approach. In these two states, the error boundary value of the core error loop is used as the position error boundary and the velocity error boundary. When the dot product is less than zero and the relative distance is not greater than the warning distance threshold, the two machines are in a state of rapid approach, and the error boundary value of the limit error loop is switched to be used as the position error boundary and the velocity error boundary.

[0010] Preferably, the construction of the relative collision cone region using the dynamic horizontal radius of the dynamic airspace protection zone as a safety interval includes: An expanded circular protection zone is constructed with the fusion location of the intruder as the center and the dynamic horizontal radius as the radius; Two tangent lines are drawn from the fusion location to the expansion circle protection zone, and the area between the two tangent lines forms a relative collision cone region; Calculate the collision cone half-angle based on the dynamic horizontal radius and the relative distance between the local machine and the intruding machine; The relative velocity vector is obtained based on the difference between the fusion speed of the local machine and the fusion speed of the intrusion machine, and the relative velocity deflection angle between the relative velocity vector and the line connecting the positions of the two machines is calculated. When the relative velocity deflection angle is less than the half angle of the collision cone, it is determined that the relative velocity vector is within the relative collision cone area; otherwise, it is determined to be a safe state and the alarm process of the current sampling period is terminated.

[0011] Preferably, if the relative velocity vector of the machine is not in the relative collision cone region, it is determined to be safe; otherwise, the time for the machine to enter and exit the dynamic airspace protection zone is calculated, and it is determined whether there is an airspace conflict and the conflict entry time is obtained, including: Based on the horizontal relative position and horizontal relative velocity in the fused state vector, and combined with the dynamic horizontal radius, the horizontal entry time and horizontal exit time are calculated. Based on the vertical relative position and vertical relative velocity in the fused state vector, and combined with the dynamic vertical half height, the vertical entry time and vertical exit time are calculated. The larger of the horizontal entry time and the vertical entry time is taken as the conflict entry time, and the smaller of the horizontal exit time and the vertical exit time is taken as the conflict exit time. When the conflict entry time is less than the conflict exit time, the conflict entry time is within a preset prediction time window, and the horizontal distance condition and the vertical distance condition are met, it is determined that there is a clearance conflict.

[0012] Preferably, when the alarm level is determined to be recent, the heading and ground speed conflict boundary values ​​are calculated based on the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone. The conflict is then determined for the divided detection zones in conjunction with the vertical boundary conditions. Zones with conflicts are assigned to the warning zone, and zones without conflicts are assigned to the safe maneuver zone. This includes: Based on the fused state vector, the turning radius and turning angular velocity of the machine's heading turn maneuver are determined, and the nonlinear relative motion trajectory equation formed by the machine's turn is established; the common tangent between the nonlinear relative motion trajectory and the protection circle with the dynamic horizontal radius as the radius is solved by Cauchy inequality to obtain the heading conflict boundary value, which includes the upper tangent boundary angle and the lower tangent boundary angle; Based on the fused state vector, the acceleration of the machine's variable speed maneuver is determined, and the equation of the nonlinear relative motion trajectory formed by the machine's variable speed is established; the common tangent between the nonlinear relative motion trajectory of the variable speed and the protection circle is solved by Cauchy inequality to obtain the ground speed conflict boundary value, which includes acceleration boundary value and deceleration boundary value. Based on the dynamic vertical half-height and the vertical relative position and vertical relative velocity in the fused state vector, calculate the vertical entry time and vertical exit time of entering and leaving the dynamic airspace protection zone in the vertical direction; at the maneuvering time corresponding to the vertical entry time and the vertical exit time, solve the nonlinear relative motion trajectory equations simultaneously to obtain the additional heading boundary value and the additional ground speed boundary value. The heading conflict boundary values ​​and the additional heading boundary values ​​are arranged in ascending order of angle, and a heading detection interval is formed between adjacent boundary values; The ground speed conflict boundary value and the additional ground speed boundary value are arranged in ascending order of speed, and a ground speed detection interval is formed between adjacent boundary values; For each of the heading detection intervals and each of the ground speed detection intervals, the median angle theorem is used. When there is a clearance conflict in the maneuver state corresponding to the median of the interval, it is determined that all maneuver values ​​in the corresponding detection interval are in conflict and are assigned to the warning zone. Otherwise, the corresponding detection interval is assigned to the safe maneuver zone.

[0013] Preferably, after determining that the alarm level is recent, the method further includes: When the entire heading detection zone or the ground speed detection zone is included in the warning zone, the conflict prevention zone is determined to be saturated, and the recovery phase is triggered. The machine obtains the start time of the first entry into the dynamic airspace protection zone and the end time of the first exit from the dynamic airspace protection zone during the horizontal turn maneuver. The heading angle corresponding to the end time of the maneuver is used as the recommended heading angle for recovery, and the ground speed value corresponding to the end time of the maneuver is used as the recommended ground speed value for recovery. Calculate the difference between the recommended recovery heading angle and the current machine heading angle for left and right turns respectively, select the direction with the smaller difference as the recommended recovery maneuver direction, and classify the maneuver range corresponding to the recommended recovery maneuver direction into the recovery maneuver zone; Calculate the difference between the recommended recovery ground speed value and the current machine ground speed value for acceleration and deceleration respectively, select the direction with the smaller difference as the recommended recovery speed direction, and classify the speed range corresponding to the recommended recovery speed direction into the recovery maneuver zone.

[0014] Secondly, this application discloses a conflict alarm device for manned-unmanned aerial vehicle (UAV) integrated operation, which adopts the following technical solution, including: The performance error module is used to acquire multi-source data from the local machine and the intrusion machine based on the sampling period, and fuse them through Kalman filtering to obtain the fused state vector and covariance matrix; based on the covariance matrix, it constructs performance error components for the manned and unmanned ends, and synthesizes each performance error component into a core error loop and a limit error loop, which together constitute the error safety margin; The airspace protection module is used to construct a basic airspace protection zone centered on the fusion position of the intruder based on the fusion state vector; and to dynamically expand the horizontal radius and vertical half-height of the basic protection zone using the error safety margin to obtain a dynamic airspace protection zone. The conflict determination module is used to construct a relative collision cone region with the dynamic horizontal radius of the dynamic airspace protection zone as the safety interval; if the relative velocity vector of the machine is not in the relative collision cone region, it is determined to be safe; otherwise, the time for the machine to enter and exit the dynamic airspace protection zone is calculated to determine whether there is an airspace conflict and to obtain the conflict entry time. The alarm level module is used to determine the alarm level based on the conflict entry time and a preset threshold range. The alarm levels include: recent alarm level, mid-term warning level and security level. The boundary delineation module, when determined to be at a recent alarm level, calculates the heading and ground speed conflict boundary values ​​based on the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone. It then combines the vertical boundary conditions to determine the conflict of the divided detection zones. Zones with conflicts are assigned to the warning zone, while zones without conflicts are assigned to the safe maneuver zone.

[0015] Thirdly, this application also provides a control device, the device comprising: It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed, such as the conflict alarm method for manned-unmanned aircraft fusion operation described above.

[0016] Fourthly, this application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above for a conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation.

[0017] In summary, this application acquires multi-source surveillance data from both the local and intrusion aircraft in each sampling period. This data is then fused using Kalman filtering to obtain a fused state vector and covariance matrix. Based on the covariance matrix, performance error components for both the manned and unmanned aircraft are constructed and synthesized into a core error loop and a limit error loop, forming an error safety margin. Next, a basic airspace protection zone is constructed centered on the intrusion aircraft's fused location. The horizontal radius and vertical half-height of the protection zone are dynamically expanded using the error safety margin, and the zone is adaptively switched between the core and limit error loops based on the relative motion of the two aircraft, resulting in a dynamic airspace protection zone. A relative collision cone region is constructed using the dynamic horizontal radius as a safety interval for the first-level rapid pre-screening. If the relative velocity vector is outside the collision cone, safety is directly determined; otherwise, a second-level fine detection is initiated, calculating the time to enter and exit the dynamic protection zone and determining if an airspace conflict exists to obtain the conflict entry time. Finally, the conflict entry time is compared with an adaptive alarm time threshold, and a graded alarm result (recent alarm, mid-term warning, or safety level) is output. Finally, when the warning level is determined to be near-term, the Cauchy inequality is used to solve for the common tangent between the aircraft's maneuver trajectory and the protection circle of the dynamic protection zone, directly obtaining the heading and ground speed conflict boundary values. Combined with the vertical boundary conditions, the mean angle theorem is used to determine the conflict in each detection interval, generating a visual conflict prevention zone consisting of warning zones and safety maneuver zones. Thus, through error safety margin-driven and two-level progressive detection, highly reliable graded conflict warnings are achieved under sensor error conditions, providing pilots with specific and intuitive maneuver avoidance guidance. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a conflict alarm method for manned and unmanned aerial vehicle (UAV) integrated operations.

[0019] Figure 2 This is a structural block diagram of a conflict warning device that integrates manned and unmanned aerial vehicle (UAV) operations. Detailed Implementation

[0020] The following combination Figures 1-2 This application will be described in further detail.

[0021] Reference Figure 1 The embodiments of this application include at least steps S10 to S50.

[0022] S10 acquires multi-source data from the local machine and the intrusion machine based on the sampling period, and fuses them through Kalman filtering to obtain the fused state vector and covariance matrix; constructs performance error components for the manned and unmanned ends based on the covariance matrix, and synthesizes each performance error component into a core error loop and a limit error loop, which together constitute the error safety margin.

[0023] S20, based on the fusion state vector, constructs a basic airspace protection zone with the fusion position of the intruder as the center; and dynamically expands the horizontal radius and vertical half-height of the basic protection zone using the error safety margin to obtain a dynamic airspace protection zone.

[0024] S30: Construct a relative collision cone region with the dynamic horizontal radius of the dynamic airspace protection zone as the safety interval; if the relative velocity vector of the machine is not in the relative collision cone region, it is considered safe; otherwise, calculate the time for the machine to enter and exit the dynamic airspace protection zone, determine whether there is an airspace conflict and obtain the conflict entry time.

[0025] S40 determines the alarm level based on the conflict entry time and a preset threshold range. The alarm levels include: recent alarm level, mid-term warning level, and security level.

[0026] S50, when the alarm level is determined to be recent, the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone are used to calculate the conflict boundary value of the heading and ground speed. Combined with the vertical boundary conditions, the conflict is determined for the divided detection zone. The zone with conflict is assigned to the warning zone, and the zone without conflict is assigned to the safe maneuver zone.

[0027] Specifically, in each sampling period, multi-source data such as ADS-B and satellite navigation from both the local and intruding aircraft are fused. Kalman filtering is used to obtain a fused state vector containing position and velocity, and a covariance matrix representing data uncertainty. From this, performance error components for communication, navigation, and surveillance are extracted, constructing an error safety margin consisting of a core error loop and a limit error loop. The former covers normal error fluctuations, while the latter reserves extreme error boundaries. Next, a standard cylindrical airspace protection zone is established centered on the intruding aircraft's position. The position and velocity errors from the error safety margin are weighted by a time variable and superimposed onto the horizontal radius and vertical half-height of the protection zone. The system adaptively switches between the two error loops based on the relative motion states of the two aircraft, such as moving away, slowly approaching, or rapidly approaching, forming a dynamic airspace protection zone. Conflict detection employs a two-stage progressive approach: first, relative collision cones based on the velocity barrier method are constructed at dynamic horizontal radius intervals for rapid pre-screening of obviously conflict-free scenarios; after pre-screening, the DAITALUS airspace logic precisely calculates the entry and exit times of the dynamic protection zone to determine whether an airspace conflict has occurred and extracts the conflict entry time. The alarm decision-making process compares the conflict entry time with an adaptive alarm threshold, outputting a classification result of recent alarm, mid-term warning, or safety level. When the error switches to the limit error loop, the threshold is automatically adjusted upwards to provide early warning. For recent alarms, the Cauchy inequality is used to directly and analytically solve for the common tangent between the aircraft's turning or speed-changing maneuver trajectory and the dynamic protection zone, obtaining the heading and ground speed conflict boundary values ​​in one step. Combined with the vertical boundary conditions, the median angle theorem is used to determine the conflict state of each detection interval, ultimately generating a visualized conflict prevention zone composed of warning zones and safety maneuver zones. This method achieves highly reliable graded conflict alarms even under the condition of sensor measurement errors by driving the dynamic expansion of the protection zone through full-process penetration of error safety margin, balancing real-time performance and accuracy with two-level progressive detection, and accelerating boundary solution by replacing iterative traversal with analytical methods. It also provides specific and intuitive maneuver avoidance guidance for remote pilots.

[0028] In some embodiments, during each sampling period, multi-source data from both the local machine and the intruder are collected via airborne ADS-B transceivers, satellite navigation receivers, airborne radar or photoelectric sensors, C2 data links, and weather sensors. These data sources differ in data type, update frequency, and error characteristics: ADS-B data has a high update frequency but depends on the target machine's broadcast signal; satellite navigation data is highly autonomous but affected by ionospheric delay; and airborne sensor data has high accuracy but limited operating range.

[0029] To fuse the aforementioned multi-source data, this embodiment uses position and velocity information from ADS-B broadcast data as state prediction input, and position and velocity information from satellite navigation data and airborne sensor detection data as observation update input, employing Kalman filtering for recursive fusion. Let the sampling period number be k, and the state vector... Includes the position and velocity components of both the local and intruding machines in the local East-North-Sky coordinate system. The state prediction equation is: ; The observation update equation is: ; in, Here is the state transition matrix. Here is the Kalman gain matrix. For the observation vector, This is the observation matrix. Let represent the predicted state vector at time k based on the state at time k-1. This represents the state vector updated by observation at time k-1. This represents the state vector updated after observation at time k. The fused state vector obtained after fusion includes the local fused position. Local fusion speed Intrusion machine fusion location Integration speed with intrusion machines Simultaneously, the covariance matrix representing the uncertainty of the data is obtained. .

[0030] Location variance is extracted from the submatrix corresponding to the local machine in the covariance matrix. , and velocity variance , Combined with wind speed variance in meteorological observation data The performance error components of the UAV are constructed. Among them, the satellite navigation and positioning error is taken as the root mean square value of the position variance. ; Wind speed yaw error is based on stochastic differential equations: Modeling, .

[0031] The total error at the drone end is: ; Based on the submatrix corresponding to the intruder in the covariance matrix, the position variance and velocity variance are extracted. Combined with preset communication performance accuracy parameters c, navigation performance accuracy parameters n, and surveillance performance refresh rate parameters s, a manned-machine performance error component is constructed. (Communication performance positioning error) Navigation performance and positioning error Monitoring performance positioning error Total error at the manned end for: ; The core error loop and the limit error loop combine the root mean square values ​​of all error components from both the UAV and manned aircraft ends to obtain the error boundary values ​​of the core error loop. This error loop covers conventional error fluctuations, and the actual flight state falls within it with a preset confidence probability. The result is the sum of the error boundary value of the core error loop and the maximum value of each error component multiplied by a preset amplification factor k. The error boundary values ​​of the limit error loop are obtained. This error loop reserves extreme error boundaries to capture the worst-case scenario at the tail of the error distribution. The aforementioned core error loop and limit error loop nested together constitute the error safety margin.

[0032] In some embodiments, the dynamic airspace protection zone is constructed centered on the intruder's fusion location after obtaining an error safety margin. A preset minimum horizontal separation threshold is used. For the radius, preset the minimum vertical separation threshold. A cylindrical protected area is constructed to accommodate half-height objects. Based on this, a preset time threshold is introduced. The cylindrical protection zone is horizontally extended forward along the direction of relative motion between the two aircraft to form a basic clearance protection zone. The mathematical definition of the basic clearance protection zone is: ... if and only if ... , and At that time, both aircraft were in a state of airspace intrusion, among which, To correct for the closest point time, The distance to the nearest point. This refers to the vertically relative position.

[0033] ; when The time indicates the time it takes for the two aircraft to reach the nearest boundary of the protected area after considering the radius of the protected area; ; ; The time to reach the point closest to the horizontal. This represents the relative position components of the local machine and the intruder in the vertical direction within the fused state vector.

[0034] To achieve dynamic expansion of the protected area, the location error boundary is extracted from the error safety margin. and velocity error boundary Calculate dynamic time variables ,in The estimated time to reach the nearest boundary of the protected area, based on the fused state vector and error safety margin correction: ; in, Relative position This refers to relative velocity.

[0035] By adding the error terms to the horizontal radius and vertical half-height respectively, we obtain the dynamic horizontal radius and dynamic vertical half-height: ; ; In some embodiments, considering the relative motion state determination and error loop switching logic, the corresponding processing steps are as follows: The determination of the relative motion state of the two machines and its coupling logic with the error loop switching are as follows: Calculate the dot product of the relative velocity vector and the relative position vector of the two machines. .when At that time, the two aircraft were far apart; when and At this time, the two machines are in a slow approaching state. In both states, the error boundary values ​​of the core error loop are used as the position error boundary and velocity error boundary. This dual-loop adaptive switching mechanism ensures that the expansion of the protected area matches the actual level of risk.

[0036] Collision detection employs a two-stage progressive approach. The first stage is a rapid pre-screening based on an improved velocity barrier method: using the fusion location of the intruder as the center and a dynamic horizontal radius... An expansion circle protection zone is constructed with radius [missing information]. Two tangents are drawn from the fusion location of the host machine to this expansion circle protection zone, and the area between the two tangents forms a relative collision cone region. The collision cone half-angle [missing information]. and relative velocity deflection angle The calculation formula is: ; It should be noted that, It is the position vector pointing from the local machine to the intruder, and its direction is relative to the aforementioned position. Conversely, the vector direction pointing from the local machine to the intruder is used here to maintain geometric consistency when calculating the angle with the relative velocity vector.

[0037] when If the relative velocity vector is within the relative collision cone region, it is determined that there is a potential horizontal conflict, and the second level of fine detection is initiated; otherwise, it is determined to be a safe state and the alarm process of the current sampling cycle is terminated.

[0038] The second level is fine-grained collision detection based on DAIDALUS headroom logic. Based on the relative motion state, the entry and exit times in the horizontal and vertical directions are calculated separately. Horizontal entry time... and horizontal departure time satisfy: ; Vertical entry time and vertical departure time satisfy: ; Take the larger entry time and the smaller exit time: ; hour, If, within the preset prediction time window, and if the horizontal and vertical distance conditions are met, an airspace conflict is determined, and the conflict initiation time is obtained. .

[0039] In the alarm decision-making process, the conflict entry time is compared with the adaptive alarm threshold. A first alarm time threshold is preset. Second alarm time threshold .

[0040] when At that time, it was determined to be a recent warning level, and an immediate collision avoidance maneuver was instructed. When the protected area expansion adopts a limit error loop, the alarm threshold is automatically increased. ; When the alarm level is determined to be recent, the collision prevention zone generation process begins. Assume the machine operates at a constant turning angular velocity. and turning radius Perform a horizontal turning maneuver. Taking a right turn as an example, the coordinates of any point on the turning trajectory of the machine in the absolute motion coordinate system are: ; ; in, This indicates the initial heading angle of the machine at the current sampling time. This represents the angle that the machine has rotated through over time t since the current moment. , The x and y components are the fused positions of the local machine. , Let x and y be the coordinates of any point on the turning trajectory of the machine.

[0041] Considering the intruding aircraft flying at a constant speed in a straight line, the equations of the nonlinear relative trajectories of the two aircraft are: ; ; in, , The x and y coordinates of the relative motion trajectories of the two aircraft at time t are the coordinates of the turning trajectory of the aircraft minus the coordinates of the straight flight trajectory of the intruding aircraft. , The x and y components represent the fusion location of the intrusion machine; , Let x and y be the components of the intrusion machine's fusion velocity.

[0042] To solve for this nonlinear relative motion trajectory and the dynamic horizontal radius Let the common tangent between the protective circles of radius be denoted by Cauchy's inequality. Let the relative trajectories be... The tangent point at time is The slope of the tangent at that point is The equation of the tangent is In this equation, coefficients A and B are determined by the derivatives of the relative trajectory at the point of tangency, representing the normal vector components of the tangent in the x and y directions, respectively, which together determine the direction of the tangent; C is the intercept term of the tangent equation, determined by the coordinates of the point of tangency and the slope of the tangent, representing the position of the tangent in the coordinate plane.

[0043] From the two-dimensional form of Cauchy's inequality: ; in, To preserve the coordinates of the circle's center, we substitute the intercept and rearrange to obtain the equation for the common tangent: ; This single equation transforms the incremental iterative process of traversing all heading angles one by one in the original DAITALUS formal model into directly solving an analytical equation. Solving this equation yields the tangent time. and the moment of slicing down Substituting the turning trajectory, the upper tangent boundary angle in the heading conflict boundary value can be obtained. and the incised boundary angle Similarly, based on the nonlinear relative trajectory formed by the machine's variable-speed maneuver with a fixed acceleration *a*, the same Cauchy inequality method is used to solve for the common tangent, thus obtaining the acceleration boundary value in the ground speed conflict boundary value. and deceleration boundary value .

[0044] In some embodiments, vertical boundary conditions also need to be considered in terms of three-dimensional spatial integrity. This is based on dynamic vertical half-height. and the vertical relative position in the fusion state vector and vertical relative velocity Calculate vertical entry time and vertical departure time : ; Will and Substituting the corresponding maneuver moments into the nonlinear relative motion trajectory equations and solving them simultaneously yields the additional heading boundary values. , and additional ground speed boundary value , .

[0045] In some embodiments, the heading conflict boundary values ​​and additional heading boundary values ​​are arranged in ascending order of angle, forming heading detection intervals between adjacent boundary values; the ground speed conflict boundary values ​​and additional ground speed boundary values ​​are arranged in ascending order of speed, forming ground speed detection intervals between adjacent boundary values. The median value of each detection interval is taken, and the median angle theorem is used for conflict determination. The median angle theorem guarantees that within an interval formed by a given set of boundary values, if the maneuvering state corresponding to the median angle of the interval conflicts, then all other maneuvering angles within that interval also conflict, and the conflict states are continuous. When the maneuvering state corresponding to the median value of the interval has a clearance conflict, it is determined that all maneuvering values ​​within that detection interval conflict and are classified into the warning zone; otherwise, the detection interval is classified into the safe maneuvering zone. The warning zone and the safe maneuvering zone together constitute the conflict prevention zone, presented in visual forms such as red and green on the ground control station interface, allowing remote pilots to intuitively select the safe maneuvering range in both horizontal heading and horizontal ground speed dimensions.

[0046] In some embodiments, when the entire heading detection zone or ground speed detection zone is included in the warning zone, the conflict prevention zone is determined to be saturated, indicating that both aircraft have entered the cylindrical protection zone, triggering the recovery phase processing. This embodiment employs a direct target value decoupling strategy instead of the traditional bisection iterative search. The specific steps are as follows: Obtain the start time of the maneuver when the aircraft first enters the dynamic airspace protection zone during a horizontal turn maneuver. and the moment of termination of the maneuver when first leaving the dynamic airspace protection zone . The moment when the vehicle first points to a safe direction outside the protected area during a turning maneuver, its corresponding heading angle is determined. Recommended heading angle for recovery (subscript L indicates that this angle is derived from the disengagement time) (D indicates a safe direction pointing outside the protected area), and its corresponding ground velocity value. Recommended ground speed value for recovery: .

[0047] The differences between the recommended recovery heading angle for left and right turns and the current aircraft heading angle, as well as the differences between the recommended recovery ground speed for acceleration and deceleration and the current aircraft ground speed, are calculated separately. The direction with the smaller difference is selected as the recommended maneuver direction, and the maneuver range corresponding to the recommended maneuver direction is included in the recovery maneuver zone. The recovery maneuver zone, along with the warning zone and the safety maneuver zone, are visualized and output together to provide pilots with complete multi-level maneuver guidance.

[0048] The implementation principle of a conflict alarm method for manned-unmanned aircraft and unmanned aircraft fusion operation according to an embodiment of this application is as follows: In each sampling period, multi-source monitoring data from the local aircraft and the intruding aircraft are acquired. These data are then fused using Kalman filtering to obtain a fusion state vector and a covariance matrix. Based on the covariance matrix, performance error components for both the manned and unmanned aircraft are constructed and synthesized into a core error loop and a limit error loop, which together constitute an error safety margin. Next, a basic airspace protection zone is constructed centered on the fusion location of the intruding aircraft. The horizontal radius and vertical half-height of the protection zone are dynamically expanded using the error safety margin. The zone is adaptively switched between the core error loop and the limit error loop based on the relative motion state of the two aircraft, resulting in a dynamic airspace protection zone. A relative collision cone region is constructed using the dynamic horizontal radius as a safety interval for the first-level rapid pre-screening. If the relative velocity vector is outside the collision cone, safety is directly determined; otherwise, a second-level fine detection is performed, calculating the time to enter and exit the dynamic protection zone and determining whether an airspace conflict exists to obtain the conflict entry time. Then, the conflict entry time is compared with an adaptive alarm time threshold, and a graded alarm result (recent alarm, mid-term warning, or safety level) is output. Finally, when a near-term alarm level is determined, the Cauchy inequality is used to solve for the common tangent between the aircraft's maneuver trajectory and the protection circle of the dynamic protection zone, directly obtaining the heading and ground speed conflict boundary values. Combined with the vertical boundary conditions, the median angle theorem is used to determine conflicts in each detection interval, generating a visualized conflict prevention zone composed of warning zones and safety maneuver zones. The double-ring nested structure of the error safety margin in this application, and its full-process coupling drive with the dynamic expansion of the protection zone and alarm threshold adjustment, solve the problem that traditional methods with fixed thresholds cannot adapt to error fluctuations. The two-level progressive architecture of rapid pre-screening using the speed barrier method and fine detection using the DAITALUS airspace logic balances the real-time performance and accuracy of conflict detection. The analytical solution of the common tangent using the Cauchy inequality replaces the conventional incremental traversal iteration, significantly reducing computational load while improving detection accuracy. The graded alarm combined with adaptive threshold adjustment ensures early warning when sensor errors increase and reduces false alarms when errors decrease. Ultimately, under the condition of sensor measurement errors, a highly reliable graded conflict alarm is achieved, providing specific maneuver avoidance guidance for remote pilots.

[0049] Figure 1 This is a flowchart illustrating a conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation in one embodiment. It should be understood that, although... Figure 1The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless explicitly stated otherwise, there is no strict order requirement for the execution of these steps, and they can be executed in other orders; and Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0050] Based on the same technical concept, referring to Figure 2 This application also provides a conflict alarm device for manned-unmanned aerial vehicle (UAV) integrated operation, which adopts the following technical solution: The device includes: The performance error module is used to acquire multi-source data from the local machine and the intrusion machine based on the sampling period, and fuse them through Kalman filtering to obtain the fused state vector and covariance matrix; based on the covariance matrix, it constructs performance error components for the manned and unmanned ends, and synthesizes each performance error component into a core error loop and a limit error loop, which together constitute the error safety margin; The airspace protection module is used to construct a basic airspace protection zone based on the fusion state vector and centered on the fusion position of the intruder; and to dynamically expand the horizontal radius and vertical half-height of the basic protection zone using the error safety margin to obtain a dynamic airspace protection zone. The conflict determination module is used to construct a relative collision cone region with the dynamic horizontal radius of the dynamic airspace protection zone as the safety interval. If the relative velocity vector of the machine is not in the relative collision cone region, it is determined to be safe. Otherwise, the time for the machine to enter and exit the dynamic airspace protection zone is calculated to determine whether there is an airspace conflict and to obtain the conflict entry time. The alarm level module is used to determine the alarm level based on the conflict entry time and a preset threshold range. The alarm levels include: recent alarm level, medium-term warning level, and security level. The boundary delineation module is used to calculate the heading and ground speed conflict boundary values ​​based on the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone when the recent alarm level is determined. Combined with the vertical boundary conditions, it determines the conflict of the divided detection zones. Zones with conflicts are assigned to the warning zone, and zones without conflicts are assigned to the safe maneuver zone.

[0051] In some embodiments, the performance error module is specifically used to utilize Kalman filtering, taking the position and velocity information in ADS-B broadcast data as the state prediction input, and taking the position and velocity information in satellite navigation data and airborne sensor detection data as the observation update input to perform recursive fusion, so as to obtain a fused state vector and a covariance matrix. The fused state vector includes the local fused position, the local fused velocity, the intruder fused position, and the intruder fused velocity. Based on the submatrix corresponding to the local machine in the covariance matrix, the position variance and velocity variance are extracted. Combined with the wind speed variance in meteorological observation data, the UAV end performance error components are constructed. The UAV end performance error components include satellite navigation positioning error and wind speed yaw error. Based on the submatrix corresponding to the intruder in the covariance matrix, the position variance and velocity variance are extracted. Combined with the preset communication performance accuracy parameters, navigation performance accuracy parameters and surveillance performance refresh rate parameters, the manned aircraft performance error components are constructed. The manned aircraft performance error components include communication performance positioning error, navigation performance positioning error and surveillance performance positioning error. Based on the error values ​​of each component in the performance error of the UAV and the performance error of the manned aircraft, the error boundary value of the core error loop is obtained; based on the error boundary value of the core error loop, the maximum value of each error component, and the preset amplification factor, the error boundary value of the limit error loop is obtained.

[0052] In some embodiments, the airspace protection module is specifically used to construct a cylindrical protection zone with the intruder fusion location as the center, a preset horizontal minimum separation threshold as the radius, and a preset vertical minimum separation threshold as the half height; Based on a preset time threshold, the cylindrical protection zone is horizontally extended forward along the relative motion direction of the two aircraft to form a basic airspace protection zone. The position error boundary and velocity error boundary are extracted from the error safety margin; the dynamic time variable is calculated, which is the smaller value between the preset prediction time window and the corrected closest point time. The corrected closest point time is the estimated time to reach the nearest boundary of the protected area after correction based on the fused state vector and the error safety margin. The dynamic horizontal radius is obtained by superimposing the horizontal radius with the position error boundary and the sum of the dynamic time variable multiplied by the velocity error boundary; the dynamic vertical half-height is obtained by superimposing the vertical position error boundary with the dynamic time variable multiplied by the vertical velocity error boundary, thus forming the dynamic airspace protection zone. Specifically, the dot product of the relative velocity vector and the relative position vector of the two machines is calculated. When the dot product is not less than zero, the two machines are in a state of distance. When the dot product is less than zero and the relative distance between the two machines is greater than a preset warning distance threshold, the two machines are in a state of slow approach. In these two states, the error boundary value of the core error loop is used as the position error boundary and the velocity error boundary. When the dot product is less than zero and the relative distance is not greater than the warning distance threshold, the two machines are in a state of rapid approach, and the error boundary value of the limit error loop is switched as the position error boundary and the velocity error boundary.

[0053] In some embodiments, the conflict determination module is specifically used to construct an expansion circle protection zone with the fusion location of the intruder as the center and the dynamic horizontal radius as the radius; Two tangents are drawn from the fusion location to the expansion circle protection zone, and the area between the two tangents forms a relative collision cone region; The collision cone half-angle is calculated based on the dynamic horizontal radius and the relative distance between the local machine and the intruding machine; The relative velocity vector is obtained based on the difference between the fusion speed of the local machine and the fusion speed of the intrusion machine. The relative velocity deflection angle between the relative velocity vector and the line connecting the positions of the two machines is then calculated. When the relative velocity deflection angle is less than half the collision cone angle, the relative velocity vector is determined to be within the relative collision cone region; otherwise, it is determined to be a safe state and the alarm process of the current sampling period is terminated.

[0054] In some embodiments, the conflict determination module is specifically used to calculate the horizontal entry time and horizontal exit time based on the horizontal relative position and horizontal relative velocity in the fused state vector, combined with the dynamic horizontal radius. Based on the vertical relative position and vertical relative velocity in the fused state vector, combined with the dynamic vertical half height, the vertical entry time and vertical exit time are calculated. The larger of the horizontal entry time and the vertical entry time is taken as the conflict entry time, and the smaller of the horizontal exit time and the vertical exit time is taken as the conflict exit time. An airspace conflict is determined to exist when the conflict entry time is less than the conflict exit time, the conflict entry time is within the preset prediction time window, and the horizontal and vertical distance conditions are met.

[0055] In some embodiments, the boundary delineation module is specifically used to determine the turning radius and turning angular velocity of the machine's heading turn maneuver based on the fused state vector, establish the nonlinear relative motion trajectory equation formed by the machine's turn, and use Cauchy's inequality to solve for the common tangent between the nonlinear relative motion trajectory and the protection circle with the dynamic horizontal radius as the radius, to obtain the heading conflict boundary value, which includes the upper tangent boundary angle and the lower tangent boundary angle. The acceleration of the machine's variable speed maneuver is determined based on the fused state vector, and the equation of the nonlinear relative motion trajectory formed by the machine's variable speed is established. The common tangent between the nonlinear relative motion trajectory of the variable speed and the protection circle is solved by Cauchy inequality to obtain the ground speed conflict boundary value, which includes acceleration boundary value and deceleration boundary value. Based on the vertical relative position and vertical relative velocity in the dynamic vertical half height and fused state vector, the vertical entry time and vertical exit time of entering and leaving the dynamic airspace protection zone in the vertical direction are calculated; at the maneuvering time corresponding to the vertical entry time and vertical exit time, the nonlinear relative motion trajectory equations are solved simultaneously to obtain the additional heading boundary value and the additional ground speed boundary value. Arrange the heading conflict boundary values ​​and additional heading boundary values ​​in ascending order of angle, and form a heading detection interval between adjacent boundary values; Arrange the conflicting ground speed boundary values ​​and the additional ground speed boundary values ​​in ascending order of speed, and form a ground speed detection interval between adjacent boundary values; For each heading detection zone and each ground speed detection zone, the median value of the zone is taken. Using the median angle theorem, when there is a clearance conflict in the maneuver state corresponding to the median value of the zone, it is determined that all maneuver values ​​in the corresponding detection zone are in conflict and are assigned to the warning zone; otherwise, the corresponding detection zone is assigned to the safe maneuver zone.

[0056] In some embodiments, the boundary delineation module is also used to determine that the conflict prevention zone is saturated and trigger the recovery phase processing when the entire heading detection zone or ground speed detection zone is classified into the warning zone. The machine obtains the start time of the first entry into the dynamic airspace protection zone and the end time of the first exit from the dynamic airspace protection zone during the horizontal turn maneuver. The heading angle corresponding to the end time of the maneuver is used as the recommended heading angle for recovery, and the ground speed value corresponding to the end time of the maneuver is used as the recommended ground speed value for recovery. Calculate the difference between the recommended recovery heading angle for left turn and right turn and the current heading angle of the machine, select the direction with the smaller difference as the recommended recovery maneuver direction, and classify the maneuver range corresponding to the recommended recovery maneuver direction into the recovery maneuver zone; Calculate the difference between the recommended recovery ground speed value and the current machine ground speed value for acceleration and deceleration respectively. Select the direction with the smaller difference as the recommended recovery speed direction and classify the speed range corresponding to the recommended recovery speed direction into the recovery maneuver zone.

[0057] This application also discloses a control device.

[0058] Specifically, the control device includes a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed to perform the aforementioned conflict alarm method for manned-unmanned aerial vehicle (UAV) fusion operation.

[0059] This application also discloses a computer-readable storage medium.

[0060] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed, such as the conflict alarm method for the integrated operation of manned and unmanned aircraft described above. The computer-readable storage medium includes, for example, various media that can store program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0061] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation, characterized in that, include: Based on the sampling period, multi-source data from the local machine and the intrusion machine are acquired and fused using Kalman filtering to obtain the fused state vector and covariance matrix; Based on the covariance matrix, performance error components for manned and unmanned aircraft are constructed, and each performance error component is combined into a core error loop and a limit error loop, which together constitute the error safety margin. Based on the fusion state vector, a basic airspace protection zone is constructed with the fusion location of the intruder as the center. The horizontal radius and vertical half-height of the basic protection zone are dynamically expanded using the aforementioned error safety margin to obtain a dynamic clearance protection zone. Using the dynamic horizontal radius of the dynamic airspace protection zone as a safety interval, a relative collision cone region is constructed; if the relative velocity vector of the machine is not within the relative collision cone region, it is considered safe. Otherwise, calculate the time for the machine to enter and exit the dynamic airspace protection zone, determine whether there is an airspace conflict, and obtain the conflict entry time; Based on the conflict entry time and a preset threshold range, the alarm level is determined, including: recent alarm level, mid-term warning level, and security level; When the alarm level is determined to be recent, the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone are used to calculate the conflict boundary value of the heading and ground speed. Combined with the vertical boundary conditions, the conflict of the divided detection zones is determined. Zones with conflicts are assigned to the warning zone, and zones without conflicts are assigned to the safe maneuver zone.

2. The conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation according to claim 1, characterized in that, The multi-source data includes: ADS-B broadcast data, satellite navigation data, airborne sensor detection data, C2 link data, and meteorological observation data; The process involves constructing performance error components for both manned and unmanned aircraft based on the covariance matrix, and then synthesizing these performance error components into a core error loop and a limit error loop, which together constitute the error safety margin, including: Using the Kalman filter, position and velocity information from ADS-B broadcast data are used as state prediction input, and position and velocity information from satellite navigation data and airborne sensor detection data are used as observation update input for recursive fusion to obtain the fused state vector and the covariance matrix. The fused state vector includes the local fused position, local fused velocity, intruder fused position, and intruder fused velocity. Based on the submatrix corresponding to the local machine in the covariance matrix, the position variance and velocity variance are extracted. Combined with the wind speed variance in the meteorological observation data, the UAV end performance error components are constructed. The UAV end performance error components include satellite navigation positioning error and wind speed yaw error. Based on the submatrix corresponding to the intruder in the covariance matrix, the position variance and velocity variance are extracted. Combined with the preset communication performance accuracy parameters, navigation performance accuracy parameters and monitoring performance refresh rate parameters, the manned aircraft performance error components are constructed. The manned aircraft performance error components include communication performance positioning error, navigation performance positioning error and monitoring performance positioning error. Based on the error values ​​of the UAV-side performance error components and the manned aircraft-side performance error components, the error boundary value of the core error loop is obtained; based on the error boundary value of the core error loop, the maximum value of each error component, and a preset amplification factor, the error boundary value of the limit error loop is obtained.

3. The conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation according to claim 2, characterized in that, The basic airspace protection zone is constructed based on the fusion state vector, with the fusion position of the intruder as the center; The horizontal radius and vertical half-height of the basic protection zone are dynamically expanded using the aforementioned error safety margin to obtain a dynamic clearance protection zone, including: A cylindrical protection zone is constructed with the intrusion machine fusion location as the center, a preset horizontal minimum separation threshold as the radius, and a preset vertical minimum separation threshold as the half height. Based on a preset time threshold, the cylindrical protection zone is horizontally extended forward along the relative motion direction of the two aircraft to form the basic airspace protection zone. The position error boundary and velocity error boundary are extracted from the error safety margin; the dynamic time variable is calculated, which is the smaller value between the preset prediction time window and the corrected closest point time. The corrected closest point time is the estimated time to reach the nearest boundary of the protected area after correction based on the fused state vector and the error safety margin. The dynamic horizontal radius is obtained by superimposing the horizontal radius with the position error boundary and the sum of the dynamic time variable multiplied by the velocity error boundary; the dynamic vertical half-height is obtained by superimposing the vertical position error boundary with the dynamic time variable multiplied by the vertical velocity error boundary, thus forming the dynamic airspace protection zone. Specifically, the dot product of the relative velocity vector and the relative position vector of the two machines is calculated. When the dot product is not less than zero, the two machines are in a state of distance. When the dot product is less than zero and the relative distance between the two machines is greater than a preset warning distance threshold, the two machines are in a state of slow approach. In these two states, the error boundary value of the core error loop is used as the position error boundary and the velocity error boundary. When the dot product is less than zero and the relative distance is not greater than the warning distance threshold, the two machines are in a state of rapid approach, and the error boundary value of the limit error loop is switched to be used as the position error boundary and the velocity error boundary.

4. The conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation according to claim 3, characterized in that, The construction of a relative collision cone region, using the dynamic horizontal radius of the dynamic airspace protection zone as a safety interval, includes: An expanded circular protection zone is constructed with the fusion location of the intruder as the center and the dynamic horizontal radius as the radius; Two tangent lines are drawn from the fusion location to the expansion circle protection zone, and the area between the two tangent lines forms a relative collision cone region; Calculate the collision cone half-angle based on the dynamic horizontal radius and the relative distance between the local machine and the intruding machine; The relative velocity vector is obtained based on the difference between the fusion speed of the local machine and the fusion speed of the intrusion machine, and the relative velocity deflection angle between the relative velocity vector and the line connecting the positions of the two machines is calculated. When the relative velocity deflection angle is less than the half angle of the collision cone, it is determined that the relative velocity vector is within the relative collision cone area; otherwise, it is determined to be a safe state and the alarm process of the current sampling period is terminated.

5. A conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation according to claim 4, characterized in that, If the relative velocity vector of the machine is not in the relative collision cone area, then it is considered safe; Otherwise, calculate the time for the machine to enter and exit the dynamic airspace protection zone, determine whether an airspace conflict exists and obtain the conflict entry time, including: Based on the horizontal relative position and horizontal relative velocity in the fused state vector, and combined with the dynamic horizontal radius, the horizontal entry time and horizontal exit time are calculated. Based on the vertical relative position and vertical relative velocity in the fused state vector, and combined with the dynamic vertical half height, the vertical entry time and vertical exit time are calculated. The larger of the horizontal entry time and the vertical entry time is taken as the conflict entry time, and the smaller of the horizontal exit time and the vertical exit time is taken as the conflict exit time. When the conflict entry time is less than the conflict exit time, the conflict entry time is within a preset prediction time window, and the horizontal distance condition and the vertical distance condition are met, it is determined that there is a clearance conflict.

6. A conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation according to claim 1, characterized in that, When a recent alarm level is determined, based on the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone, the heading and ground speed conflict boundary values ​​are calculated. Combined with the vertical boundary conditions, conflict is determined for the divided detection zones. Zones with conflicts are assigned to the warning zone, and zones without conflicts are assigned to the safe maneuver zone, including: Based on the fused state vector, the turning radius and turning angular velocity of the machine's heading turn maneuver are determined, and the nonlinear relative motion trajectory equation formed by the machine's turn is established; the common tangent between the nonlinear relative motion trajectory and the protection circle with the dynamic horizontal radius as the radius is solved by Cauchy inequality to obtain the heading conflict boundary value, which includes the upper tangent boundary angle and the lower tangent boundary angle; Based on the fused state vector, the acceleration of the machine's variable speed maneuver is determined, and the equation of the nonlinear relative motion trajectory formed by the machine's variable speed is established; the common tangent between the nonlinear relative motion trajectory of the variable speed and the protection circle is solved by Cauchy inequality to obtain the ground speed conflict boundary value, which includes acceleration boundary value and deceleration boundary value. Based on the dynamic vertical half-height and the vertical relative position and vertical relative velocity in the fused state vector, calculate the vertical entry time and vertical exit time of entering and leaving the dynamic airspace protection zone in the vertical direction; at the maneuvering time corresponding to the vertical entry time and the vertical exit time, solve the nonlinear relative motion trajectory equations simultaneously to obtain the additional heading boundary value and the additional ground speed boundary value. The heading conflict boundary values ​​and the additional heading boundary values ​​are arranged in ascending order of angle, and a heading detection interval is formed between adjacent boundary values; The ground speed conflict boundary value and the additional ground speed boundary value are arranged in ascending order of speed, and a ground speed detection interval is formed between adjacent boundary values; For each of the heading detection intervals and each of the ground speed detection intervals, the median angle theorem is used. When there is a clearance conflict in the maneuver state corresponding to the median of the interval, it is determined that all maneuver values ​​in the corresponding detection interval are in conflict and are assigned to the warning zone. Otherwise, the corresponding detection interval is assigned to the safe maneuver zone.

7. A conflict alarm method for manned-unmanned aerial vehicle (UAV) integrated operation according to claim 6, characterized in that, After determining the alarm level as recent, the following is also included: When the entire heading detection zone or the ground speed detection zone is included in the warning zone, the conflict prevention zone is determined to be saturated, and the recovery phase is triggered. The machine obtains the start time of the first entry into the dynamic airspace protection zone and the end time of the first exit from the dynamic airspace protection zone during the horizontal turn maneuver. The heading angle corresponding to the end time of the maneuver is used as the recommended heading angle for recovery, and the ground speed value corresponding to the end time of the maneuver is used as the recommended ground speed value for recovery. Calculate the difference between the recommended recovery heading angle and the current machine heading angle for left and right turns respectively, select the direction with the smaller difference as the recommended recovery maneuver direction, and classify the maneuver range corresponding to the recommended recovery maneuver direction into the recovery maneuver zone; Calculate the difference between the recommended recovery ground speed value and the current machine ground speed value for acceleration and deceleration respectively, select the direction with the smaller difference as the recommended recovery speed direction, and classify the speed range corresponding to the recommended recovery speed direction into the recovery maneuver zone.

8. A conflict alarm device for manned-unmanned aerial vehicle (UAV) integrated operation, characterized in that, The device includes: The performance error module is used to acquire multi-source data from the local machine and the intrusion machine based on the sampling period, and fuse them through Kalman filtering to obtain the fused state vector and covariance matrix; based on the covariance matrix, it constructs performance error components for the manned and unmanned ends, and synthesizes each performance error component into a core error loop and a limit error loop, which together constitute the error safety margin; The airspace protection module is used to construct a basic airspace protection zone centered on the fusion position of the intruder based on the fusion state vector; and to dynamically expand the horizontal radius and vertical half-height of the basic protection zone using the error safety margin to obtain a dynamic airspace protection zone. The conflict determination module is used to construct a relative collision cone region with the dynamic horizontal radius of the dynamic airspace protection zone as the safety interval; if the relative velocity vector of the machine is not in the relative collision cone region, it is determined to be safe; otherwise, the time for the machine to enter and exit the dynamic airspace protection zone is calculated to determine whether there is an airspace conflict and to obtain the conflict entry time. The alarm level module is used to determine the alarm level based on the conflict entry time and a preset threshold range. The alarm levels include: recent alarm level, mid-term warning level, and security level. The boundary delineation module, when determined to be at a recent alarm level, calculates the heading and ground speed conflict boundary values ​​based on the nonlinear relative motion trajectory formed by the machine's maneuver and the horizontal protection radius of the dynamic airspace protection zone. It then combines the vertical boundary conditions to determine the conflict of the divided detection zones. Zones with conflicts are assigned to the warning zone, while zones without conflicts are assigned to the safe maneuver zone.

9. A control device, characterized in that, The device includes: A memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1 to 7.

Images