Low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking

Abstract

The application relates to the technical field of low-altitude airspace control, and discloses a low-altitude anti-collision linkage method based on airspace geometric deconstruction and two-dimensional projection tracking; the method establishes a space constraint pool containing a permission type space constraint object and a restriction type space constraint object, and obtains a current available passing space through three-dimensional space difference set operation; according to a heading deviation angle and a convergence trend of a measured carrier, an environmental disturbance deviation state or a high-risk continuous invasion state is determined, and a local boundary expansion or an alarm level adjustment is correspondingly executed; based on a three-dimensional motion state of the measured carrier, air resistance parameters and a local wind field vector, a set of touch-down point probabilities is calculated, the set of touch-down point probabilities is projected to form a two-dimensional falling risk area, and a final risk level is determined in combination with a ground dynamic target area; risk information is directionally sent to a corresponding ground interaction node according to a risk state and a space position, and linkage early warning of low-altitude abnormal risks and ground targets is realized.

Description

Technical Field

[0001] This invention relates to the field of low-altitude airspace management technology, specifically to a low-altitude collision avoidance linkage method based on airspace geometric deconstruction and two-dimensional projection tracking. Background Technology

[0002] With the increasing application of low-altitude test vehicles such as drones and eVTOL in urban logistics, inspection, and emergency response scenarios, urban low-altitude airspace exhibits a coexistence of various spatial constraints, including static three-dimensional entities such as buildings, commercial logistics flight channels, and temporary restricted areas. Current low-altitude airspace management typically uses pre-defined electronic fences, fixed routes, or no-fly zones to constrain test vehicles, combining real-time location and heading information for boundary violation judgment and alerts. However, when different types and priority levels of airspace constraints overlap in time and space, existing methods often lack the ability to dynamically reconstruct the overlapping areas, easily resorting to overall closure or uniform avoidance, making it difficult to fully utilize remaining passable space while ensuring high-priority control needs. Furthermore, for short-term deviations caused by environmental factors such as wind disturbances, simply alerting based on location boundary violations can easily confuse regressive environmental disturbance deviations with abnormal behavior continuously moving towards restricted areas, leading to false alarms or alarm levels that do not match the actual risk.

[0003] Furthermore, existing low-altitude safety monitoring typically focuses on the flight path compliance and collision avoidance of the monitored vehicle, but lacks effective means to analyze the potential ground fall risk caused by dynamic anomalies in the monitored vehicle in conjunction with dynamic ground targets. Directly using the complete three-dimensional motion process for continuous calculation and multi-terminal display not only involves a large amount of three-dimensional spatial data and calculations, but also hinders ground personnel from quickly identifying potentially affected areas. Summary of the Invention

[0004] The purpose of this invention is to provide a low-altitude collision avoidance linkage method based on airspace geometric deconstruction and two-dimensional projection tracking, which solves the technical problems in the existing low-altitude airspace management process, such as the difficulty in dynamically determining the actual passable space when spatial constraints of different attributes and priority levels overlap in time and space, the difficulty in distinguishing between short-term deviations caused by environmental disturbances and high-risk states of continuous approach to restricted space, and the difficulty in associating and directionally linking the potential crash area after an aircraft power anomaly with dynamic ground targets for handling.

[0005] The objective of this invention can be achieved through the following technical solutions: A low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking, the method comprising the following steps: S1: Establish a spatial constraint pool, input permitted and restricted spatial constraint objects, and record the effective time period, spatial boundary, and control priority level of each object; obtain the real-time three-dimensional position, current velocity vector, motion state, air resistance parameters, and local wind field vector of the tested vehicle. S2: Filter the currently valid spatial constraint objects according to the valid time period. Based on the candidate passage space limited by the permission type object, perform three-dimensional spatial difference operation on the restricted objects that overlap with the candidate passage space to obtain the currently available passage space. S3: Determine the heading deviation angle of the tested vehicle relative to the planned route based on the current velocity vector, and determine whether there is a convergence trend towards reverting to the planned route based on the changes in the heading deviation angle over multiple consecutive sampling periods; the convergence trend refers to the absolute value of the heading deviation angle in the current sampling period being less than the average value of the previous few sampling periods, and the absolute value of the heading deviation angle showing a continuous decreasing trend over multiple consecutive sampling periods; when the tested vehicle exceeds the currently available passage space and has the aforementioned convergence trend, it is determined that the tested vehicle is in an environmental disturbance deviation state, and the local boundary of the currently available passage space is expanded outward by a preset safety distance, and the initial alarm level is determined; S4: Based on the real-time three-dimensional location, the motion state, the air resistance parameters, and the local wind field vector, calculate the set of ground contact point probabilities under abnormal dynamic conditions, project it onto a two-dimensional coordinate system to generate a two-dimensional fall risk area; obtain the ground dynamic target area, and perform a two-dimensional spatial intersection judgment between the two-dimensional fall risk area and the ground dynamic target area, and determine the final risk level based on the intersection judgment result; S5: Based on the risk status and corresponding alarm level of the tested vehicle, and in conjunction with the real-time three-dimensional location, determine the target ground interaction node and send the corresponding risk information to it.

[0006] As a further aspect of the present invention: in step S2, obtaining the currently available passage space includes: Based on the business attributes of each spatial constraint object, spatial constraint objects are divided into permissive spatial constraint objects and restricted spatial constraint objects; Filter out restricted spatial constraint objects that overlap with the permitted spatial constraint objects in both the effective time period and the three-dimensional spatial range; Using the candidate passage space corresponding to the permissive spatial constraint object as the subtracted space and the currently valid restrictive spatial constraint object as the clipping space, perform a three-dimensional spatial difference operation to obtain the currently available passage space.

[0007] As a further aspect of the present invention: in step S3, the determination of the environmental disturbance deviation state and the expansion of the local boundary include: Calculate the heading deviation angle between the current speed direction of the tested vehicle and the tangential direction of the planned route; When the tested vehicle exceeds the currently available passage space, and the absolute value of the heading deviation angle in multiple consecutive sampling periods meets the convergence criterion, the tested vehicle is determined to be in an environmental disturbance deviation state. The convergence criterion refers to the following: the absolute value of the heading deviation angle in the current sampling period is less than the arithmetic mean of the absolute values ​​of the heading deviation angle in the previous several sampling periods, and the absolute value of the heading deviation angle shows a continuous downward trend in multiple consecutive sampling periods. The predicted offset direction is determined based on the local wind field vector and the motion state of the measured vehicle. Without intruding on the currently effective restricted space constraints, the boundary located on one side of the predicted offset direction in the currently available passage space is expanded locally. The distance of the local boundary expansion is a preset safety distance. Correction guidance information is generated, and the current alarm level is set to a preset low alarm level.

[0008] As a further aspect of the present invention: step S3 further includes: When the extension line of the current velocity direction of the tested vehicle intersects geometrically with a high-priority restricted spatial constraint object, and the heading deviation angle does not meet the convergence determination condition, the tested vehicle is determined to be in a high-risk continuous intrusion state. In response to the high-risk continuous intrusion state, the expansion of the local boundary of the currently available passage space is prohibited, and the current alarm level is raised to the preset highest alarm level.

[0009] As a further aspect of the present invention: in step S4, the generation of the two-dimensional fall risk area includes: The system acquires the initial altitude, initial velocity, and air resistance parameters of the tested vehicle when it is in a state of dynamic anomaly, and also acquires the local wind field vector in the current airspace. The initial height, initial velocity, air resistance parameters, gravity parameters, and local wind field vector are input into a preset dynamic anomaly fall motion model, and the set of ground contact point probabilities is obtained through multiple iterative calculations. The set of ground contact point probabilities is projected onto a two-dimensional coordinate system, and a two-dimensional fall risk zone is generated based on the distribution range of the two-dimensional ground contact points.

[0010] As a further aspect of the present invention: in step S4, the two-dimensional spatial intersection determination includes: The system acquires ground dynamic data representing dense crowds or congested traffic flow from the city's Internet of Things (IoT) platform, and generates ground dynamic target areas based on the ground dynamic data. The two-dimensional fall risk area and the dynamic target area on the ground are transformed to the same two-dimensional coordinate system, and the overlapping area and overlapping area of ​​the two are calculated. When the overlapping area is greater than the preset area tolerance threshold, it is determined that there is a risk of ground impact, and the final risk level is determined to be the preset highest risk level.

[0011] As a further aspect of the present invention: In step S5, when the tested vehicle is in an environmental disturbance deviation state, data including the initial alarm level, two-dimensional location information and correction guidance information is sent to the operation scheduling terminal corresponding to the tested vehicle.

[0012] As a further aspect of the present invention: In step S5, when the final risk level is the highest risk level, the core evacuation zone is determined based on the overlapping area of ​​the two-dimensional fall risk area and the ground dynamic target area, and the data containing the final risk level and the location information of the core evacuation zone is sent to the on-site handling terminal that meets the preset distance condition from the core evacuation zone.

[0013] As a further aspect of the present invention: the permitted space constraint objects include authorized commercial logistics flight channels, and the restricted space constraint objects include fixed obstacle spaces corresponding to urban buildings and terrain, as well as temporary restricted spaces designated due to emergencies; wherein, a corresponding priority level is set according to the control attributes of the restricted space constraint objects.

[0014] The beneficial effects of this invention are: (1) This invention distinguishes between permissive spatial constraint objects and restricted spatial constraint objects, and uses three-dimensional spatial difference operations to eliminate restricted spatial regions in candidate passage spaces to obtain the currently available passage space. This solves the technical problem of lack of dynamic geometric reconstruction capability when different types of spatial constraints overlap in time and space in the prior art, thereby reducing the loss of available space caused by the overall closed candidate passage space.

[0015] (2) The present invention makes state judgment based on the change of the heading deviation angle of the tested carrier and its convergence trend towards the planned route, and performs local boundary expansion and generates correction guidance information for environmental disturbance deviation state. The distance of the local boundary expansion is a preset safety distance, and the alarm level is increased for high-risk continuous intrusion state, thereby realizing differentiated processing of different deviation states.

[0016] (3) This invention uses the initial height, initial speed, air resistance parameters and local wind field vector of the tested vehicle to predict the probability set of the ground landing point and form a two-dimensional fall risk area. By making spatial intersection judgment with the dynamic target area on the ground, the abnormal state in the air and the potential affected target on the ground are spatially correlated.

[0017] (4) The present invention determines the target ground interaction node based on the risk status and corresponding location information of the tested carrier, and sends the corresponding risk information to the operation scheduling terminal or the on-site disposal terminal to provide data support for track correction, on-site warning and ground evacuation disposal. Attached Figure Description

[0018] The invention will now be further described with reference to the accompanying drawings.

[0019] Figure 1 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 As shown, in one embodiment, a low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking is provided. The method includes the following steps: S1: Establish a spatial constraint pool, input permitted and restricted spatial constraint objects, and record the effective time period, spatial boundary, and control priority level of each object; obtain the real-time three-dimensional position, current velocity vector, motion state, air resistance parameters, and local wind field vector of the tested vehicle. S2: Filter the currently valid spatial constraint objects according to the valid time period. Based on the candidate passage space limited by the permission type object, perform three-dimensional spatial difference operation on the restricted objects that overlap with the candidate passage space to obtain the currently available passage space. S3: Determine the heading deviation angle of the tested vehicle relative to the planned route based on the current velocity vector, and determine whether there is a convergence trend towards reverting to the planned route based on the changes in the heading deviation angle over multiple consecutive sampling periods; the convergence trend refers to the absolute value of the heading deviation angle in the current sampling period being less than the average value of the previous few sampling periods, and the absolute value of the heading deviation angle showing a continuous decreasing trend over multiple consecutive sampling periods; when the tested vehicle exceeds the currently available passage space and has the aforementioned convergence trend, it is determined that the tested vehicle is in an environmental disturbance deviation state, and the local boundary of the currently available passage space is expanded outward by a preset safety distance, and the initial alarm level is determined; S4: Based on the real-time three-dimensional location, the motion state, the air resistance parameters, and the local wind field vector, calculate the set of ground contact point probabilities under abnormal dynamic conditions, project it onto a two-dimensional coordinate system to generate a two-dimensional fall risk area; obtain the ground dynamic target area, and perform a two-dimensional spatial intersection judgment between the two-dimensional fall risk area and the ground dynamic target area, and determine the final risk level based on the intersection judgment result; S5: Based on the risk status and corresponding alarm level of the tested vehicle, and in conjunction with the real-time three-dimensional location, determine the target ground interaction node and send the corresponding risk information to it.

[0022] Specifically, this method runs on a backend server with data processing capabilities. The specific steps are as follows: In step S1, the server initializes the data structure of the spatial constraint pool, parses the received permissive and restricted spatial constraint objects into data containing three-dimensional spatial boundary coordinates and valid time periods, and writes them into the spatial constraint pool.

[0023] In this embodiment, spatial constraint objects are categorized into permitted spatial constraint objects and restricted spatial constraint objects based on their effect on the access rights of the tested vehicle. Permitted spatial constraint objects limit the three-dimensional spatial range that the tested vehicle is authorized to traverse, such as an approved commercial logistics flight corridor. Restricted spatial constraint objects limit the three-dimensional spatial range that the tested vehicle is prohibited from entering or is subject to restricted entry into, such as fixed obstacle spaces corresponding to urban buildings and temporary restricted areas formed by emergencies. For restricted spatial constraint objects, corresponding priority levels are set according to their control attributes.

[0024] In step S2, the server filters the currently valid spatial constraint objects according to the valid time period corresponding to each spatial constraint object, and determines whether the restricted spatial constraint object overlaps with the candidate passage space corresponding to the permitted spatial constraint object. For the restricted spatial constraint objects that overlap, a three-dimensional spatial difference operation is performed with the candidate passage space as the subtracted object to obtain the three-dimensional boundary data of the currently available passage space.

[0025] In step S3, the server acquires motion state data of the tested vehicle over multiple consecutive sampling periods and calculates the heading deviation angle based on the current velocity vector and the tangential vector of the current position of the planned route. When the three-dimensional position of the tested vehicle exceeds the boundary of the currently available passage space, the server further judges the change in the absolute value of the heading deviation angle over multiple consecutive sampling periods. When the absolute value of the heading deviation angle generally shows a decreasing trend, it is determined that the tested vehicle has a convergence trend towards reverting to the planned route, and this state is judged as an environmental disturbance deviation state. In response to the environmental disturbance deviation state, the server determines the predicted offset direction based on the local wind field vector and the motion state of the tested vehicle. Without intruding on the currently effective restricted space constraints, the server performs local boundary expansion on the boundary located on one side of the predicted offset direction in the currently available passage space. The distance of the local boundary expansion is a preset safety distance, and corresponding initial alarm level and correction guidance information are generated.

[0026] In step S4, the server first retains the height information of the real-time three-dimensional position of the tested vehicle, and calculates the fall process under abnormal dynamic conditions by combining the motion state of the tested vehicle, air resistance parameters and local wind field vector, and obtains the set of ground contact probability. Then, the set of ground contact probability is projected onto a two-dimensional coordinate system to generate a two-dimensional fall risk area. The two-dimensional fall risk area is spatially intersected with the dynamic target area on the ground, and the final risk level is determined based on the judgment result.

[0027] In step S5, the server reads the preset role distribution rules and determines the target ground interaction node based on the risk status, initial alarm level or final risk level of the tested vehicle and the corresponding location information. Subsequently, the corresponding alarm level, risk level, two-dimensional location information and / or correction guidance information are encapsulated into risk data and sent to the corresponding operation dispatch terminal or on-site handling terminal through the preset communication interface.

[0028] Through the above technical solution, this embodiment obtains the currently available passage space by performing three-dimensional spatial calculations on the permitted and restricted spatial constraint objects, and distinguishes and processes different deviation states by combining the convergence trend of the heading deviation angle; at the same time, it predicts the set of ground contact probability based on the three-dimensional motion state of the tested vehicle, air resistance parameters, and local wind field vectors, and maps it to form a two-dimensional fall risk area, which is then correlated with the dynamic target area on the ground. Based on the risk status, risk information is sent to the corresponding operation dispatch terminal or on-site disposal terminal, realizing the data connection between airspace status judgment, ground risk analysis, and coordinated disposal.

[0029] It should be noted that the "expansion of the local boundary within the defined range" refers to shifting the selected currently available passage space boundary outward by a preset safe distance, provided that it does not intrude upon the currently valid restricted spatial constraints. The safe distance can be a fixed value depending on the type of the tested vehicle, for example, 10 meters for a UAV and 20 meters for an electric vertical takeoff and landing (EVTOL) aircraft; or it can be dynamically determined based on the current local wind field intensity, for example, 1.2 times the maximum lateral drift that the tested vehicle may experience under the current wind speed, but not exceeding the preset maximum expansion distance limit, such as 50 meters.

[0030] In one embodiment of the present invention, in step S2, obtaining the currently available passage space includes: Based on the business attributes of each spatial constraint object, spatial constraint objects are divided into permissive spatial constraint objects and restricted spatial constraint objects; Filter out restricted spatial constraint objects that overlap with the permitted spatial constraint objects in both the effective time period and the three-dimensional spatial range; Using the candidate passage space corresponding to the permissive spatial constraint object as the subtracted space and the currently valid restrictive spatial constraint object as the clipping space, perform a three-dimensional spatial difference operation to obtain the currently available passage space.

[0031] Specifically, firstly, the server divides the spatial constraint objects into permitted spatial constraint objects and restricted spatial constraint objects according to the business attributes of each spatial constraint object, and then filters the currently valid spatial constraint objects according to the valid time period corresponding to each spatial constraint object.

[0032] Secondly, the server reads the 3D contour data of the permitted spatial constraint object and the currently valid restricted spatial constraint object, and calls the bounding box collision detection algorithm to perform an intersection test on the two in order to determine the restricted spatial constraint object that overlaps with the candidate passage space in volume.

[0033] Finally, when there is volume overlap between the candidate passage space and the restricted space constraint object, the server calls the 3D graphics Boolean operator to perform a 3D spatial difference operation with the candidate passage space corresponding to the permitted space constraint object as the subtracted object and the currently valid restricted space constraint object as the clipping object. This process removes the space portion in the candidate passage space that overlaps with the restricted space constraint object and recalculates the vertex connections along the cutting section to form a closed 3D topology, thus obtaining the currently available passage space.

[0034] Through the above technical solution, this embodiment distinguishes between permitted and restricted spatial constraints. Based on the candidate passage space formed by the permitted spatial constraints, it uses a three-dimensional spatial difference operation to eliminate spatial regions that overlap with currently valid restricted spatial constraints, thereby obtaining the currently usable passage space. Thus, while satisfying the requirements of restricted spatial constraints, it retains the unrestricted remaining areas in the candidate passage space, reducing the loss of usable space caused by completely closing the candidate passage space and improving the utilization efficiency of low-altitude space.

[0035] In one embodiment of the present invention, step S3 includes determining the environmental disturbance deviation state and expanding the local boundary: Calculate the heading deviation angle between the current speed direction of the tested vehicle and the tangential direction of the planned route; When the tested vehicle exceeds the currently available passage space, and the absolute value of the heading deviation angle in multiple consecutive sampling periods meets the convergence criterion, the tested vehicle is determined to be in an environmental disturbance deviation state. The convergence criterion refers to the following: the absolute value of the heading deviation angle in the current sampling period is less than the arithmetic mean of the absolute values ​​of the heading deviation angle in the previous several sampling periods, and the absolute value of the heading deviation angle shows a continuous downward trend in multiple consecutive sampling periods. The predicted offset direction is determined based on the local wind field vector and the motion state of the measured vehicle. Without intruding on the currently effective restricted space constraints, the boundary located on one side of the predicted offset direction in the currently available passage space is expanded locally. The distance of the local boundary expansion is a preset safety distance. Correction guidance information is generated, and the current alarm level is set to a preset low alarm level.

[0036] Specifically, firstly, environmental disturbance deviation status is determined: the server acquires motion state data of the tested vehicle over multiple consecutive sampling periods, extracts the current velocity vector, and calculates the heading deviation angle between the current velocity direction and the tangential direction of the planned route. When it is confirmed through three-dimensional coordinate comparison that the tested vehicle exceeds the boundary of the currently available passage space, the server monitors the change in the heading deviation angle over multiple consecutive sampling periods; if the absolute value of the heading deviation angle generally shows a decreasing trend, it is determined that the tested vehicle has a convergence trend towards reverting to the planned route, and this state is identified as an environmental disturbance deviation state.

[0037] Let the sampling period be N, the current sampling period be k, and the absolute value of the heading deviation angle be |θ|. k The values ​​of |θ| for the first N-1 sampling periods are respectively |θ| k-1 |, ..., |θ k-N+1 |

[0038] The judgment condition is: the current value is less than the average value of the previous N-1 periods, that is... Furthermore, the linear regression slope of |θ| is less than 0 over the most recent M consecutive periods. Here, N ranges from 3 to 10, and M ranges from 5 to 20.

[0039] Secondly, local boundary expansion and guidance are performed: In response to the above-mentioned environmental disturbance deviation, the server obtains the local wind field vector through the meteorological data interface, and determines the predicted offset direction based on the vector synthesis direction of the local wind field vector and the current velocity vector of the measured vehicle.

[0040] Without intruding on the currently effective restricted spatial constraints, an outward coordinate offset is applied to the boundary vertices located on one side of the predicted offset direction in the currently available passage space to form a local boundary expansion. The distance of the local boundary expansion is a preset safety distance: when the tested vehicle is an unmanned aerial vehicle (UAV), the safety distance is 10 meters; when the tested vehicle is an electric vertical take-off and landing (EVTOL) aircraft, the safety distance is 20 meters; or, the safety distance is dynamically determined based on the current local wind field intensity, taking 1.2 times the maximum lateral drift that the tested vehicle may produce under the current wind speed, and not exceeding 50 meters.

[0041] Simultaneously, the server calculates the corrected direction from the current position of the tested vehicle to the planned route. Specifically, it generates a direction vector from the current position of the tested vehicle to the nearest point on the planned route as the corrected heading, and uses this to generate corrected guidance information to guide the tested vehicle back to the planned route. The server sets the current alarm level to the preset low alarm level.

[0042] Through the above technical solution, this embodiment determines whether the tested vehicle has a convergence trend towards returning to the planned route based on the change of the heading deviation angle within multiple consecutive sampling periods. When it is determined to be a deviation state due to environmental disturbance, the local boundary of the currently available passage space is expanded. The distance of the local boundary expansion is a preset safety distance. At the same time, correction guidance information is generated, thereby distinguishing and handling short-term deviations caused by environmental disturbances from other high-risk deviation states, reducing false alarms caused by alarms based solely on position crossing the boundary, and providing guidance information for the tested vehicle to return to the planned route.

[0043] As one embodiment of the present invention, step S3 further includes: When the extension line of the current velocity direction of the tested vehicle intersects geometrically with a high-priority restricted spatial constraint object, and the heading deviation angle does not meet the convergence determination condition, the tested vehicle is determined to be in a high-risk continuous intrusion state. In response to the high-risk continuous intrusion state, the expansion of the local boundary of the currently available passage space is prohibited, and the current alarm level is raised to the preset highest alarm level.

[0044] Specifically, firstly, a high-risk continuous intrusion state is determined: the server generates a spatial ray along its current velocity direction, starting from the current three-dimensional coordinates of the tested vehicle, and calls a ray projection algorithm to determine whether the spatial ray geometrically intersects with a high-priority restricted spatial constraint object. Simultaneously, the server continuously monitors the change in the heading deviation angle; when the spatial ray geometrically intersects with the high-priority restricted spatial constraint object, and the heading deviation angle does not show a convergence trend towards the planned route, the tested vehicle is determined to be in a high-risk continuous intrusion state.

[0045] Secondly, boundary restrictions and alarm adjustments are implemented: In response to the aforementioned high-risk continuous intrusion state, the server prohibits the expansion of the local boundary of the currently available passage space, keeping the boundary of the currently available passage space unchanged; at the same time, the current alarm level corresponding to the tested carrier is raised to the preset highest alarm level.

[0046] Through the above technical solution, this embodiment uses the geometric intersection relationship between the current speed direction of the tested vehicle and the high-priority restricted space constraint object, as well as whether the heading deviation angle has a regression convergence trend, as the basis for determining the high-risk continuous intrusion state. When it is determined to be a high-risk continuous intrusion state, the local boundary expansion of the currently available passage space is prohibited and the alarm level is increased. This distinguishes between the environmental disturbance deviation state with a regression trend and the high-risk state of continuously approaching the high-priority restricted space, thereby improving the pertinence of low-altitude abnormal state identification and alarm processing.

[0047] In one embodiment of the present invention, the generation of the two-dimensional fall risk area in step S4 includes: The system acquires the initial altitude, initial velocity, and air resistance parameters of the tested vehicle when it is in a state of dynamic anomaly, and also acquires the local wind field vector in the current airspace. The initial height, initial velocity, air resistance parameters, gravity parameters, and local wind field vector are input into a preset dynamic anomaly fall motion model, and the set of ground contact point probabilities is obtained through multiple iterative calculations. The set of ground contact point probabilities is projected onto a two-dimensional coordinate system, and a two-dimensional fall risk zone is generated based on the distribution range of the two-dimensional ground contact points.

[0048] Specifically, firstly, data preparation is performed: the server acquires the real-time three-dimensional position of the tested vehicle when it is in a dynamic abnormal state, and extracts its vertical height relative to the preset reference plane as the initial height; at the same time, it reads the air resistance parameters corresponding to the tested vehicle from the equipment attribute library, and obtains the initial speed and its direction information when the dynamic abnormal state occurs.

[0049] Secondly, fall simulation and probability generation are performed: The server obtains local wind field vector data of the current airspace through a meteorological interface, including wind speed and direction, and inputs the initial height, initial velocity, air resistance parameters, gravity parameters, and local wind field vector into a preset dynamic anomaly fall motion model. The dynamic anomaly fall motion model updates the velocity and spatial position of the tested vehicle in time steps based on gravity and the air resistance generated by the motion state of the tested vehicle relative to the local wind field, until the height of the tested vehicle reaches the preset ground height, and the corresponding position is determined as a predicted impact point; the air resistance parameters are equivalent resistance parameters pre-determined according to the type of the tested vehicle. By applying perturbation to the local wind field vector and performing multiple iterative calculations, a set of impact point probabilities consisting of multiple predicted impact points is obtained.

[0050] Finally, a two-dimensional projection and risk area generation are performed: the server projects the set of ground contact probability points onto a two-dimensional coordinate system to form a two-dimensional ground contact point distribution; based on the distribution range of the two-dimensional ground contact points, the convex hull algorithm or the alpha shape boundary extraction algorithm is used to fit the boundary of the two-dimensional ground contact point set, extract the ground contact point distribution boundary containing 80% to 95% confidence level, and the closed polygon area enclosed by the boundary is taken as the two-dimensional fall risk area.

[0051] Through the above technical solution, this embodiment uses the initial height, initial speed, air resistance parameters and local wind field vector of the tested vehicle to predict the potential contact point under abnormal dynamic conditions, and projects the contact point probability set onto a two-dimensional coordinate system to form a two-dimensional fall risk area, thereby establishing a correspondence between the three-dimensional fall process of the tested vehicle and the potential impact range on the ground, providing a data basis for subsequent two-dimensional spatial intersection judgment and risk level determination.

[0052] It should be noted that the specific iterative calculation method of the dynamic abnormal fall motion model is as follows: The time when the dynamic anomaly of the tested carrier occurs is taken as the starting time t0, and the iteration time step Δt is set, with the value of Δt ranging from 0.01 to 0.1 seconds.

[0053] The dynamic abnormal fall motion model is calculated iteratively based on the following kinematic equations: Let the three-dimensional position vector of the tested vehicle in the Cartesian coordinate system be P = (x, y, z), and the velocity vector be... The local wind field vector is .

[0054] In each iteration step Δt, the acceleration vector A of the tested vehicle is formed by the superposition of gravitational acceleration g and air resistance acceleration D, i.e., A = g + D.

[0055] in, , ρ is the gravitational constant, taken as 9.8 m / s²; The direction of D is opposite to the direction of the velocity vector V - W of the measured vehicle relative to the air, and its magnitude (i.e., the air resistance acceleration, in meters per second squared, m / s²) 2 From the formula The variables and their physical dimensions are as follows: ρ is the air density, in kilograms per cubic meter (kg / m³). 3 );C d is the drag coefficient of the tested vehicle, a dimensionless parameter; S is the reference windward area of ​​the tested vehicle, in square meters (m²). 2 ); |VW| represents the relative speed of the tested vehicle with respect to the local wind speed, in meters per second (m / s); m represents the mass of the tested vehicle, in kilograms (kg).

[0056] In each iteration cycle, update the velocity and position based on the acceleration A described above: ; ; Repeat the iteration until the height z of the tested vehicle is less than or equal to 0. At this point, record the corresponding horizontal position coordinates (x, y) as a predicted touch point.

[0057] To reflect uncertainties such as actual wind field measurement errors and atmospheric turbulence, the server applies a normally distributed random perturbation to the acquired local wind field vector. The standard deviation of the perturbation applied to wind speed is 8% to 15% of the measured wind speed value, and the standard deviation of the perturbation applied to wind direction is ±10° to ±20° of the measured wind direction angle. Under each perturbation, a complete iterative calculation of the fall process is performed independently, and a predicted impact point is recorded. By repeating the above calculation process 500 to 2000 times, a set of ground contact point probabilities consisting of multiple predicted impact points is generated for subsequent two-dimensional projection and risk area generation. The lower limit of 500 iterations ensures that the set of ground contact point probabilities has a statistically sufficient sample size to form a stable distribution, while the upper limit of 2000 iterations balances computational resource consumption with prediction accuracy, avoiding excessive iterations that could lead to system response delays.

[0058] In one embodiment of the present invention, the two-dimensional spatial intersection determination in step S4 includes: The system acquires ground dynamic data representing dense crowds or congested traffic flow from the city's Internet of Things (IoT) platform, and generates ground dynamic target areas based on the ground dynamic data. The two-dimensional fall risk area and the dynamic target area on the ground are transformed to the same two-dimensional coordinate system, and the overlapping area and overlapping area of ​​the two are calculated. When the overlapping area is greater than the preset area tolerance threshold, it is determined that there is a risk of ground impact, and the final risk level is determined to be the preset highest risk level.

[0059] Specifically, firstly, the ground dynamic data is acquired and vectorized: the server acquires real-time crowd heat maps or traffic flow density data through the data API interface of the city IoT platform, extracts the outline of connected regions where the density value reaches a preset threshold, and vectorizes the outline of the connected regions into a closed geometric figure composed of two-dimensional planar coordinates, which serves as the ground dynamic target area.

[0060] Secondly, the two-dimensional intersection area calculation is performed: the server performs discrete sampling on the boundary of the two-dimensional fall risk area and converts it into a closed polygon composed of multiple two-dimensional boundary vertices; then the closed polygon and the dynamic target area on the ground are converted to the same two-dimensional coordinate system, and the polygon Boolean intersection algorithm is called to calculate the overlapping area and overlapping region of the two.

[0061] Finally, the risk level is determined: the server compares the overlapping area with a preset area tolerance threshold; when the overlapping area is greater than the preset area tolerance threshold, it is determined that the two-dimensional fall risk area and the ground dynamic target area have effective overlap, and it is determined that there is a risk of ground fall impact, and the final risk level is determined to be the preset highest risk level.

[0062] Through the above technical solution, this embodiment makes a spatial intersection judgment between the two-dimensional fall risk area and the ground dynamic target area generated by the urban Internet of Things dynamic data, and determines the ground fall impact risk and the corresponding final risk level based on the overlapping area of ​​the two, thereby establishing a spatial association between the potential fall range of the tested vehicle and dynamic targets such as people and vehicles on the ground, providing a basis for the targeted transmission of risk information and ground handling.

[0063] It should be noted that the preset area tolerance threshold is used to exclude misjudgments caused by data noise or minimal area overlap. This threshold can be dynamically set according to the type of dynamic target area on the ground. For example, for densely populated areas, the threshold can be set to 1-5 square meters; for congested traffic areas, the threshold can be set to 10-20 square meters. In one embodiment, the threshold is set to 5 square meters by default, but it can also be configured by the system operator according to the actual scenario.

[0064] In one embodiment of the present invention, in step S5, when the tested vehicle is in an environmental disturbance deviation state, data including the initial alarm level, two-dimensional location information and correction guidance information is sent to the operation scheduling terminal corresponding to the tested vehicle.

[0065] Specifically, when the server determines that the tested vehicle is in a state of environmental disturbance deviation, it reads the device registration identifier corresponding to the tested vehicle and determines the target operation and dispatch terminal based on the preset association between the device and the operation and dispatch terminal. Subsequently, the server obtains the initial alarm level corresponding to the tested vehicle, projects its real-time three-dimensional position information onto a two-dimensional coordinate system to obtain two-dimensional position information, and reads the correction guidance information generated in step S3. The server encapsulates the initial alarm level, two-dimensional position information, and correction guidance information into risk data, and sends it to the target operation and dispatch terminal through a preset communication interface, so that the operation and dispatch terminal can prompt or dispatch the track return of the tested vehicle according to the correction guidance information.

[0066] Through the above technical solution, this embodiment can send the corresponding initial alarm level, two-dimensional location information and correction guidance information to the corresponding operation and dispatch terminal according to the deviation state of environmental disturbance. This enables the operation and dispatch terminal to obtain the deviation state and return guidance information of the tested vehicle in a timely manner, providing data support for track correction and operation scheduling, while reducing the reception of unnecessary alarm information by irrelevant terminals.

[0067] In one embodiment of the present invention, in step S5, when the final risk level is the highest risk level, the core evacuation zone is determined based on the overlapping area of ​​the two-dimensional fall risk area and the ground dynamic target area, and the data containing the final risk level and the location information of the core evacuation zone is sent to the on-site handling terminal that meets the preset distance condition from the core evacuation zone.

[0068] Specifically, when the server determines the final risk level to be the highest, it extracts the overlapping area between the two-dimensional fall risk area and the dynamic target area on the ground, and determines the core evacuation zone based on the boundary coordinates of the overlapping area. Subsequently, the server acquires the location information of multiple on-site response terminals around the core evacuation zone, calculates the distance between each on-site response terminal and the core evacuation zone, and filters target on-site response terminals according to preset distance conditions. The server encapsulates the final risk level, the location information of the core evacuation zone, and the corresponding alarm information into risk data, and sends it to the target on-site response terminals via a preset communication interface for on-site personnel to issue warnings, evacuate, or handle other risks.

[0069] Through the above technical solution, this embodiment can determine the core evacuation zone based on the overlap between the two-dimensional fall risk area and the dynamic target area on the ground when the final risk level reaches the highest risk level, and send the risk information to the on-site disposal terminal that meets the preset conditions at a distance from the core evacuation zone, thereby improving the pertinence of risk information transmission under high-risk events and providing data support for on-site warning and evacuation disposal.

[0070] As one embodiment of the present invention, the permitted space constraint objects include authorized commercial logistics flight channels, and the restricted space constraint objects include fixed obstacle spaces corresponding to urban buildings and terrain, as well as temporary restricted spaces designated due to emergencies; wherein, a corresponding priority level is set according to the control attributes of the restricted space constraint objects.

[0071] Specifically, the server obtains authorized commercial logistics flight channel data from the low-altitude general aviation approval database and constructs a three-dimensional tubular spatial model using the corresponding route coordinates, flight altitude range, and channel width as permission-type spatial constraint objects. Simultaneously, it obtains three-dimensional model data of urban buildings and terrain from the city information model platform, constructs the corresponding spatial entity ranges as fixed obstacle spaces, and obtains information such as the coordinates of the control center, control radius, and control altitude corresponding to emergencies through the emergency linkage interface to generate temporary restricted spaces. These fixed obstacle spaces and temporary restricted spaces serve as restricted spatial constraint objects. After unified coordinate processing, each spatial constraint object is written into a spatial constraint pool for subsequent spatial intersection judgments and three-dimensional spatial difference calculations between candidate passage spaces and restricted spatial constraint objects.

[0072] Through the above technical solution, this embodiment distinguishes between authorized passable spaces, fixed obstacle spaces, and temporary restricted spaces by attributes, so that spatial constraints of different natures can be uniformly modeled according to their passability attributes, providing a clear spatial data basis for the subsequent determination of currently available passable spaces.

[0073] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking, characterized in that, The method includes the following steps: S1: Establish a spatial constraint pool, input permitted and restricted spatial constraint objects, and record the effective time period, spatial boundary, and control priority level of each object; obtain the real-time three-dimensional position, current velocity vector, motion state, air resistance parameters, and local wind field vector of the tested vehicle. S2: Filter the currently valid spatial constraint objects according to the valid time period. Based on the candidate passage space limited by the permission type object, perform three-dimensional spatial difference operation on the restricted objects that overlap with the candidate passage space to obtain the currently available passage space. S3: Determine the heading deviation angle of the tested vehicle relative to the planned route based on the current velocity vector, and determine whether there is a convergence trend towards reverting to the planned route based on the changes in the heading deviation angle over multiple consecutive sampling periods; the convergence trend refers to the absolute value of the heading deviation angle in the current sampling period being less than the average value of the previous few sampling periods, and the absolute value of the heading deviation angle showing a continuous decreasing trend over multiple consecutive sampling periods; when the tested vehicle exceeds the currently available passage space and has the aforementioned convergence trend, it is determined that the tested vehicle is in an environmental disturbance deviation state, and the local boundary of the currently available passage space is expanded outward by a preset safety distance, and the initial alarm level is determined; S4: Based on the real-time three-dimensional location, the motion state, the air resistance parameters, and the local wind field vector, calculate the set of ground contact point probabilities under abnormal dynamic conditions, project it onto a two-dimensional coordinate system to generate a two-dimensional fall risk area; obtain the ground dynamic target area, and perform a two-dimensional spatial intersection judgment between the two-dimensional fall risk area and the ground dynamic target area, and determine the final risk level based on the intersection judgment result; S5: Based on the risk status and corresponding alarm level of the tested vehicle, and in conjunction with the real-time three-dimensional location, determine the target ground interaction node and send the corresponding risk information to it.

2. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 1, characterized in that, In step S2, obtaining the currently available passage space includes: Based on the business attributes of each spatial constraint object, spatial constraint objects are divided into permissive spatial constraint objects and restricted spatial constraint objects; Filter out restricted spatial constraint objects that overlap with the permitted spatial constraint objects in both the effective time period and the three-dimensional spatial range; Using the candidate passage space corresponding to the permissive spatial constraint object as the subtracted space and the currently valid restrictive spatial constraint object as the clipping space, perform a three-dimensional spatial difference operation to obtain the currently available passage space.

3. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 1, characterized in that, In step S3, the determination of the environmental disturbance deviation state and the expansion of the local boundary include: Calculate the heading deviation angle between the current speed direction of the tested vehicle and the tangential direction of the planned route; When the tested vehicle exceeds the currently available passage space, and the absolute value of the heading deviation angle in multiple consecutive sampling periods meets the convergence criterion, the tested vehicle is determined to be in an environmental disturbance deviation state. The convergence criterion refers to the following: the absolute value of the heading deviation angle in the current sampling period is less than the arithmetic mean of the absolute values ​​of the heading deviation angle in the previous several sampling periods, and the absolute value of the heading deviation angle shows a continuous downward trend in multiple consecutive sampling periods. The predicted offset direction is determined based on the local wind field vector and the motion state of the measured vehicle. Without intruding on the currently effective restricted space constraints, the boundary located on one side of the predicted offset direction in the currently available passage space is expanded locally. The distance of the local boundary expansion is a preset safety distance. Correction guidance information is generated, and the current alarm level is set to a preset low alarm level.

4. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 3, characterized in that, Step S3 also includes: When the extension line of the current velocity direction of the tested vehicle intersects geometrically with a high-priority restricted spatial constraint object, and the heading deviation angle does not meet the convergence determination condition, the tested vehicle is determined to be in a high-risk continuous intrusion state. In response to the high-risk continuous intrusion state, the expansion of the local boundary of the currently available passage space is prohibited, and the current alarm level is raised to the preset highest alarm level.

5. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 1, characterized in that, In step S4, the generation of the two-dimensional fall risk area includes: The system acquires the initial altitude, initial velocity, and air resistance parameters of the tested vehicle when it is in a state of dynamic anomaly, and also acquires the local wind field vector in the current airspace. The initial height, initial velocity, air resistance parameters, gravity parameters, and local wind field vector are input into a preset dynamic anomaly fall motion model, and the set of ground contact point probabilities is obtained through multiple iterative calculations. The set of ground contact point probabilities is projected onto a two-dimensional coordinate system, and a two-dimensional fall risk zone is generated based on the distribution range of the two-dimensional ground contact points.

6. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 5, characterized in that, In step S4, the two-dimensional spatial intersection determination includes: The system acquires ground dynamic data representing dense crowds or congested traffic flow from the city's Internet of Things (IoT) platform, and generates ground dynamic target areas based on the ground dynamic data. The two-dimensional fall risk area and the dynamic target area on the ground are transformed to the same two-dimensional coordinate system, and the overlapping area and overlapping area of ​​the two are calculated. When the overlapping area is greater than the preset area tolerance threshold, it is determined that there is a risk of ground impact, and the final risk level is determined to be the preset highest risk level.

7. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 3, characterized in that, In step S5, when the tested vehicle is in an environmental disturbance deviation state, data including the initial alarm level, two-dimensional location information and correction guidance information is sent to the operation scheduling terminal corresponding to the tested vehicle.

8. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 6, characterized in that, In step S5, when the final risk level is the highest risk level, the core evacuation zone is determined based on the overlapping area of ​​the two-dimensional fall risk area and the dynamic target area on the ground, and the data containing the final risk level and the location information of the core evacuation zone is sent to the on-site handling terminal that meets the preset distance condition from the core evacuation zone.

9. The low-altitude collision avoidance linkage method based on spatial geometry deconstruction and two-dimensional projection tracking according to claim 1, characterized in that, The permitted space constraints include authorized commercial logistics flight channels, and the restricted space constraints include fixed obstacle spaces corresponding to urban buildings and terrain, as well as temporary restricted spaces designated due to emergencies; wherein, a corresponding priority level is set according to the control attributes of the restricted space constraints.

Images