Low-altitude unmanned aerial vehicle landing path planning method and system

Abstract

This invention discloses a method and system for planning landing trajectories of low-altitude unmanned aerial vehicles (UAVs), belonging to the field of UAV trajectory planning technology. The method includes: acquiring flight status, target landing point, and predicted target aircraft trajectory; generating a main landing trajectory including the final glide segment; setting a go-around decision point on the main landing trajectory and pre-generating associated go-around trajectories; based on conflict prediction results, using the main landing trajectory as a candidate when no conflict occurs; when lateral avoidance conditions are met, adjusting the horizontal position of the reference waypoint in the conflict segment while maintaining the original planned altitude and target speed to form lateral avoidance candidates that converge back to the main landing trajectory; when go-around conditions are met, calling the associated go-around trajectories at the go-around decision point, and generating emergency go-around candidates in real time at non-go-around decision points based on the current position, current airspeed, and current mechanical energy state; performing safety verification and outputting the results. This invention can reduce the impact of avoidance on the longitudinal glide state and cover sudden go-around needs beyond fixed go-around decision points.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) trajectory planning technology, specifically relating to a method and system for planning the landing trajectory of a low-altitude UAV. Background Technology

[0002] When a low-altitude unmanned aerial vehicle (UAV) performs autonomous landing, it typically needs to generate approach and final glide paths based on the UAV's current flight status and target landing point, and handle potential traffic conflicts caused by other aircraft during flight. The waypoint altitude and target speed of the final glide phase are usually pre-planned according to the landing process. Whether the path adjustment changes these longitudinal planning states will directly affect the glide conditions on which the subsequent landing is based.

[0003] Existing landing track conflict resolution schemes may require the regeneration of a large-scale track when traffic conflicts occur during the final glide phase. This approach can easily alter the altitude or target speed of the originally planned waypoints, making it difficult for the regenerated track to maintain its original final glide state. Furthermore, if constrained by current space, maneuverability, or traffic conditions, the regenerated avoidance track may not meet the safety requirements for continued landing.

[0004] Furthermore, the need for a go-around during the final glide phase may arise either after full verification at the pre-set go-around decision point, or due to sudden severe risks at a final glide location outside the go-around decision point. If a go-around trajectory is uniformly replanned only after triggering, it adds avoidable online calculation steps at fixed decision points; if only pre-generated go-around trajectories are configured for fixed decision points, it is difficult to cover situations where immediate go-arounds are required at non-fixed locations. Therefore, a landing trajectory planning scheme is needed that can select restricted lateral avoidance or go-around based on the degree of traffic conflict, and provide go-around trajectories for both fixed and non-fixed locations. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for planning landing trajectories for low-altitude unmanned aerial vehicles (UAVs) to solve the technical problems that traffic conflict handling during the final glide phase can easily alter the original planned longitudinal glide state, and that the method for obtaining go-around trajectories cannot simultaneously adapt to both fixed go-around decision points and sudden risks at non-fixed locations.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a method for planning the landing trajectory of a low-altitude unmanned aerial vehicle (UAV), comprising:

[0008] The flight status information and target landing point information of the UAV are acquired, and the main landing trajectory of the UAV is generated. The main landing trajectory includes a final glide segment, and the final glide segment includes multiple reference waypoints. Each reference waypoint has at least a horizontal position, an altitude, and a target speed.

[0009] At least one go-around decision point is set on the main landing trajectory. A corresponding go-around trajectory is pre-generated for each go-around decision point, and the association between the go-around decision point and the corresponding go-around trajectory is established.

[0010] Obtain the predicted trajectory of the target aircraft, perform conflict prediction based on the main landing trajectory and the predicted trajectory, and determine whether a traffic conflict occurs in the final glide phase and the degree of the traffic conflict.

[0011] When it is determined that no traffic conflict has occurred based on the conflict prediction, the main landing trajectory is identified as a candidate trajectory to be executed.

[0012] When the traffic conflict occurs in the final glide segment and the degree of conflict meets the preset lateral avoidance conditions, corresponding avoidance waypoints are generated based on at least a portion of the reference waypoints in the conflict segment corresponding to the traffic conflict in the final glide segment. Each avoidance waypoint only changes its horizontal position relative to the corresponding reference waypoint, while maintaining the original planned altitude and target speed of the corresponding reference waypoint, so as to form a lateral avoidance trajectory that deviates from the main landing trajectory and then converges back to the main landing trajectory; it is determined whether the lateral avoidance trajectory meets the preset safe continued landing conditions, and if it does, the lateral avoidance trajectory is determined as the candidate trajectory to be executed;

[0013] When the traffic conflict occurs in the final glide phase and the degree of conflict meets the preset go-around conditions, or when the degree of conflict meets the preset lateral avoidance conditions but the lateral avoidance trajectory does not meet the preset safe landing conditions, a go-around is triggered and the landing along the final glide phase is stopped.

[0014] When the UAV is located at any of the go-around decision points when a go-around is triggered, the go-around trajectory associated with the go-around decision point is invoked, and the invoked go-around trajectory is determined as the candidate trajectory to be executed.

[0015] When the UAV is not at any of the go-around decision points when the go-around is triggered, the current position, current airspeed and current mechanical energy state of the UAV at the time of triggering the go-around are used as the initial reference to generate an emergency go-around trajectory in real time, and the emergency go-around trajectory is determined as the candidate trajectory to be executed.

[0016] Perform a follow-up safety check on the candidate flight path, and when the candidate flight path passes the follow-up safety check, output the candidate flight path as the flight path to be executed by the UAV.

[0017] Furthermore, generating the main landing trajectory of the UAV includes:

[0018] The wind field corresponding to the final glide path is decomposed into a longitudinal wind component along the flight direction of the final glide path and a lateral wind component perpendicular to the flight direction of the final glide path.

[0019] The longitudinal glide profile of the final glide segment is corrected according to the longitudinal wind component, a wind field compensation lateral offset is formed in the final glide segment according to the lateral wind component, and the deceleration start position of the final glide segment is adjusted according to the longitudinal wind component to generate the wind field adaptive main landing trajectory.

[0020] The wind field compensation lateral offset is used to determine the original planned horizontal position of each reference waypoint in the final glide segment.

[0021] Further, the step of performing conflict prediction based on the primary landing trajectory and the predicted trajectory, and determining the degree of the traffic conflict, includes:

[0022] Based on the main landing trajectory and the predicted trajectory, the predicted minimum separation distance and conflict time (TTC) between the UAV and the target aircraft are determined. The distance risk level is determined based on the predicted minimum separation distance, and the time risk level is determined based on the conflict time (TTC). The traffic conflict is determined to be a minor conflict, a moderate conflict, or a severe conflict based on the higher risk level between the distance risk level and the time risk level.

[0023] When the traffic conflict occurs during the final glide phase, the minor and moderate conflicts satisfy the preset lateral avoidance conditions, and the severe conflict satisfies the preset go-around conditions.

[0024] Furthermore, the mild conflict and the moderate conflict each correspond to different maximum lateral offsets, and the maximum lateral offset corresponding to the moderate conflict is greater than the maximum lateral offset corresponding to the mild conflict.

[0025] When generating the lateral avoidance trajectory, if the traffic conflict is a minor conflict or a moderate conflict, the target lateral offset is determined according to the degree of the traffic conflict, and the target lateral offset is not greater than the maximum lateral offset corresponding to the degree of conflict.

[0026] During the process of entering the conflict segment, the lateral offset of the avoidance waypoint relative to the corresponding reference waypoint is gradually increased to the target lateral offset, and after avoiding the traffic conflict, the lateral offset is gradually reduced to zero, so that the lateral avoidance track converges back to the main landing track.

[0027] Further, determining whether the lateral avoidance trajectory meets the preset safe landing conditions includes:

[0028] Verify whether the predicted minimum clearance distance between the lateral avoidance trajectory and the target aircraft meets the safety clearance requirement, whether the lateral avoidance trajectory meets the turning limit of the UAV, and whether the lateral avoidance trajectory can reconverge to the main landing trajectory;

[0029] If at least one of the verifications fails to meet the corresponding requirements, adjust the target lateral offset or the transition length of the lateral avoidance trajectory, regenerate the lateral avoidance trajectory, and re-determine whether the regenerated lateral avoidance trajectory meets the preset safe landing conditions.

[0030] Furthermore, the kinetic and potential energy of each waypoint are determined based on the airspeed and altitude of each waypoint in the main landing trajectory, and the mechanical energy state of the corresponding waypoint is determined based on the kinetic and potential energy.

[0031] For each of the aforementioned go-around decision points, the mechanical energy state at the go-around decision point is determined based on the airspeed and altitude of the main landing trajectory at that go-around decision point. The position, airspeed, and mechanical energy state of the main landing trajectory at the go-around decision point are obtained, and the position, airspeed, and mechanical energy state are used as the initial reference for the pre-generated go-around trajectory corresponding to that go-around decision point.

[0032] Furthermore, there are three go-around decision points: the glide path entrance decision point located at the entrance of the final glide segment, the glide midpoint decision point located in the middle of the final glide segment, and the decision height decision point located at a preset decision height within the final glide segment.

[0033] The glide slope inlet decision point, the glide slope midpoint decision point, and the decision height decision point correspond to the inlet energy verification point, the mid-section energy verification point, and the pre-grounding energy verification point, respectively.

[0034] The pre-generated go-around trajectory for each of the aforementioned go-around decision points includes an initial climb segment, a turning exit segment, and a waiting line segment connected in sequence; the initial climb segment is used to climb from the corresponding go-around decision point, the turning exit segment is used to turn to the waiting line direction after climbing, and the waiting line segment is used to keep the UAV in waiting flight.

[0035] Secondly, based on the same inventive concept, the present invention also provides a low-altitude unmanned aerial vehicle (UAV) landing trajectory planning system, comprising:

[0036] The data acquisition module is configured to acquire the UAV's flight status information, target landing point information, and the predicted trajectory of the target aircraft;

[0037] The main landing trajectory generation module is configured to generate the main landing trajectory of the UAV based on the flight status information and the target landing point information. The main landing trajectory includes a final glide segment, and the final glide segment includes multiple reference waypoints. Each reference waypoint has at least a horizontal position, an altitude, and a target speed.

[0038] The go-around trajectory generation module is configured to set at least one go-around decision point on the main landing trajectory, pre-generate corresponding go-around trajectories for each go-around decision point, and establish the association between the go-around decision point and the corresponding go-around trajectory.

[0039] The conflict prediction and avoidance module is configured to perform conflict prediction based on the main landing trajectory and the predicted trajectory, determine whether a traffic conflict occurs in the final glide segment and the degree of the traffic conflict, and when the traffic conflict occurs in the final glide segment and the degree of the conflict meets the preset lateral avoidance conditions, generate corresponding avoidance waypoints based on at least some reference waypoints in the conflict segment corresponding to the traffic conflict in the final glide segment. Each avoidance waypoint only changes its horizontal position relative to the corresponding reference waypoint, and maintains the original planned altitude and target speed of the corresponding reference waypoint, so as to form a lateral avoidance trajectory that deviates from the main landing trajectory and then converges back to the main landing trajectory.

[0040] The conflict prediction and avoidance module is further configured to, when it is determined that no traffic conflict has occurred based on the conflict prediction, identify the main landing trajectory as a candidate trajectory to be executed; determine whether the lateral avoidance trajectory meets the preset safe continued landing conditions, and if it does, identify the lateral avoidance trajectory as the candidate trajectory to be executed.

[0041] The conflict prediction and avoidance module is further configured to trigger a go-around and stop continuing to land along the final glide segment when the traffic conflict occurs in the final glide segment and the conflict level meets the preset go-around conditions, or when the conflict level meets the preset lateral avoidance conditions but the lateral avoidance trajectory does not meet the preset safe continuing landing conditions; when the UAV is located at any of the go-around decision points when a go-around is triggered, the go-around trajectory associated with the go-around decision point is invoked, and the invoked go-around trajectory is determined as the candidate trajectory to be executed;

[0042] The go-around trajectory generation module is further configured to, when the UAV is not at any of the go-around decision points when the go-around is triggered, use the current position, current airspeed and current mechanical energy state of the UAV at the time of triggering the go-around as the initial reference, generate an emergency go-around trajectory in real time, and determine the emergency go-around trajectory as the candidate trajectory to be executed;

[0043] The verification output module is configured to receive the candidate execution track, perform dynamic security verification on the candidate execution track, and output the candidate execution track as the execution track of the UAV when the candidate execution track passes the dynamic security verification.

[0044] Thirdly, based on the same inventive concept, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the above-described low-altitude UAV landing trajectory planning method.

[0045] Fourthly, based on the same inventive concept, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] This invention sets up three processing paths in the final glide phase traffic conflict handling: restricted lateral avoidance, fixed go-around decision point go-around, and non-fixed location emergency go-around. This allows the traffic conflict handling method to be selected according to the degree of conflict and the go-around trigger location, forming a processing chain that connects continued landing and go-around exit.

[0048] When the lateral avoidance conditions are met, the reference waypoint of the final glide segment is used as a reference. Only the horizontal position of the avoidance waypoint is changed, while the original planned altitude and target speed of the corresponding reference waypoint are maintained. This allows the lateral avoidance trajectory to converge back to the main landing trajectory after avoiding traffic conflicts, thereby reducing the disturbance of traffic avoidance on the original planned longitudinal glide state.

[0049] By assessing the safety of lateral avoidance paths for continued landing and designating them as candidate paths only when the corresponding conditions are met, the number of lateral avoidance paths that do not meet the conditions for safe continued landing can be reduced. When the conditions for a go-around are met or the lateral avoidance path ultimately fails to meet the conditions for safe continued landing, the process is switched to go-around.

[0050] By unifying the main landing trajectory, lateral avoidance trajectory, pre-generated go-around trajectory, and real-time generated emergency go-around trajectory as candidate execution trajectories, and performing follow-up safety checks before output, the trajectories formed by different processing branches are output as execution trajectories only after passing the corresponding safety checks.

[0051] When a go-around is triggered at a go-around decision point, the pre-generated go-around trajectory associated with that decision point can be invoked, which can reduce the online trajectory generation process after a go-around is triggered at a fixed decision point. When a go-around is triggered at a non-go-around decision point, an emergency go-around trajectory can be generated in real time based on the position, airspeed, and mechanical energy state at the time of the go-around trigger. This can cover situations where the final glide phase is not at a fixed position and an immediate stop landing and go-around are required.

[0052] By correcting the longitudinal glide profile and deceleration initiation position using the longitudinal wind component, and by forming a wind field compensation lateral offset using the lateral wind component, the reference waypoints for subsequent lateral avoidance already include the wind field adaptation results. By predicting the minimum separation distance and conflict time TTC respectively to determine the risk level and taking the higher level, it helps to reduce the possibility of underestimating traffic risks when only a single risk indicator is used.

[0053] By determining the mechanical energy state based on waypoint airspeed and altitude, and using the position, airspeed, and mechanical energy state at the time of the go-around trigger as the initial reference for the go-around trajectory, a basis for the state transition between the main landing trajectory and the go-around trajectory can be provided. By making the initial climb process of the go-around smoothly transition from the current descent state to the climb state, and by making the go-around acceleration process and the main landing deceleration process form an inverse relationship, it helps to reduce abrupt changes in state and energy during the go-around switchover process. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0055] Figure 2 This is a side view diagram illustrating the relationship between the two-stage main landing trajectory and wind field compensation in this invention.

[0056] Figure 3 This is a top-view diagram illustrating the relationship between the two-stage main landing trajectory and wind field compensation in this invention.

[0057] Figure 4 This is a schematic diagram of graded Bezier smoothing for small turns, medium turns, and large turns in this invention.

[0058] Figure 5 This is a schematic diagram of the final glide slope reference waypoint, avoidance waypoint, and lateral offset during the gradual entry and exit phases in this invention.

[0059] Figure 6 This is a structural block diagram of the low-altitude unmanned aerial vehicle (UAV) landing trajectory planning system of the present invention. Detailed Implementation

[0060] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0061] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this invention should have the ordinary meaning understood by those skilled in the art. The terms "first," "second," and similar terms used in the embodiments of this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0062] Example 1:

[0063] like Figure 1 As shown, this embodiment provides a method for planning landing trajectories for low-altitude unmanned aerial vehicles (UAVs). The complete processing flow includes pre-feasibility prediction, main landing trajectory generation, local smoothing of turning segments, main trajectory energy verification, pre-generation of go-around trajectories, traffic conflict handling, and follow-up safety verification. Specifically, it includes:

[0064] S100 acquires the UAV's flight status information and target landing point information, and sequentially performs range reachability prediction, terrain reachability prediction, maneuver boundary prediction, and landing weather prediction; when all four predictions are satisfied, it enters the main landing trajectory generation step; when any one prediction is not satisfied, it outputs the result that the current landing trajectory planning is not feasible.

[0065] The estimated range is based on the straight-line distance from the drone's current position to the target landing point. and cruising speed Estimate flight time. When the single-engine cruise fuel consumption rate is expressed as fuel quantity / time, the estimated minimum fuel quantity is calculated using the following formula. :

[0066]

[0067] in, This refers to the fuel consumption rate per unit during cruising. This refers to the number of engines.

[0068] when If the remaining fuel level is greater than or equal to 80%, the remaining range is deemed insufficient. For drones using other energy forms, the consumption and remaining fuel can be estimated based on the corresponding energy form, but this embodiment does not specifically limit the conversion model for other energy forms.

[0069] Terrain accessibility prediction involves extracting terrain elevation from a digital terrain database along a line connecting the UAV's current position to the target landing point. The sampling resolution of the digital terrain database is, for example, 30 meters. (Based on the starting height...) Subtract the peak value of the terrain along the line. Obtaining height difference ;when If the safety margin is less than 100m, the terrain is deemed unreachable.

[0070] Maneuver boundary prediction calculates the sum of the absolute values ​​of all heading changes in the candidate trajectory. And the total altitude change. When the total turning angle exceeds the maximum permissible total turning angle of the UAV, or the ratio of the total climb to the shortest climb time exceeds the maximum climb rate, the maneuver is deemed unreachable. Landing weather prediction compares the real-time wind speed and visibility at the target landing point with the UAV's landing weather restrictions; when the wind speed exceeds the maximum landing wind speed or the visibility is lower than the minimum landing visibility, the weather conditions are deemed unsatisfactory.

[0071] S100 generates endurance prediction results, terrain prediction results, maneuver prediction results, and weather prediction results. If any result is infeasible, it outputs the constraint type that caused the prediction to fail and the result that the current landing plan is infeasible; if all results are feasible, it provides the UAV flight status, target landing point information, wind field data, terrain data, and corresponding constraint parameters to S200. S100 is used to eliminate obviously infeasible scenarios in advance, and S200 to S700 continue to perform corresponding verifications on the actual generated flight paths.

[0072] S200: Acquire the flight status information and target landing point information of the UAV, and generate the main landing trajectory of the UAV. The main landing trajectory includes a final glide segment, and the final glide segment includes multiple reference waypoints. Each reference waypoint has at least a horizontal position, altitude, and target speed.

[0073] The main landing trajectory consists of an approach cruise phase and a final glide phase, which are connected by the glide slope entrance. The approach cruise phase extends from the UAV's current position to the glide slope entrance, primarily involving level flight or a gentle descent; the final glide phase extends from the glide slope entrance to the target landing site. The position, speed, altitude, and heading of both phases are continuous at the glide slope entrance.

[0074] The main landing track is generated in the following order: approach cruise segment waypoint generation, final glide slope longitudinal profile generation, wind field compensation, and two-segment connection verification. The position, altitude, velocity, and heading of the glide slope inlet are used as planning inputs in this embodiment, and this embodiment does not specify the algorithm for determining the inlet point. After generation, the terminal waypoint of the approach cruise segment and the first reference waypoint of the final glide slope both correspond to the glide slope inlet point, and the continuity of the two tracks is checked using this correspondence.

[0075] The approach cruise segment establishes waypoints at 1000m intervals along a line connecting the UAV's current position to the glide slope entrance, with the last waypoint located at the glide slope entrance. Each waypoint includes longitude, latitude, altitude, target speed, and heading. The longitudinal wind speed component is determined based on the angle θ between the current wind direction and the segment's heading.

[0076]

[0077] The target ground speed is determined according to the following formula:

[0078]

[0079] For wind speed, The reference airspeed is used. When there is a difference between the starting altitude and the entry point altitude, the total altitude difference is distributed to the navigation points to ensure that the average rate of climb does not exceed 10 m / s and the average rate of descent does not exceed 5 m / s.

[0080] For each waypoint, the corresponding altitude is determined based on its position within the approach cruise segment, the heading is determined based on the segment direction, and the target ground speed is determined based on the longitudinal wind speed component. The altitude change between adjacent waypoints and the flight distance are used together to calculate the average rate of climb or average rate of descent. If the calculation result exceeds the corresponding limit, the altitudes of adjacent waypoints are reallocated and re-verified; if the altitude still exceeds the corresponding limit after reallocation, the process returns to the approach cruise segment generation process to regenerate the corresponding waypoints.

[0081] For the final glide path, the wind field corresponding to the final glide path is decomposed into a longitudinal wind component along the flight direction of the final glide path and a lateral wind component perpendicular to the flight direction of the final glide path. The longitudinal glide profile of the final glide path is corrected according to the longitudinal wind component, a wind field compensation lateral offset is formed in the final glide path according to the lateral wind component, and the deceleration start position of the final glide path is adjusted according to the longitudinal wind component to generate the wind field adaptive main landing track. The wind field compensation lateral offset is used to determine the original planned horizontal position of each reference waypoint in the final glide path.

[0082] Calculate the dynamic glide angle using the following formula. :

[0083]

[0084] Use the base glide angle, for example, 3°; This represents the longitudinal wind component, with positive for headwinds and negative for tailwinds. The reference grounding point speed; This is the wind disturbance correction factor calibrated based on the aerodynamic characteristics of the UAV. Because... As a dimensionless ratio, in this embodiment... The unit of measurement is angular, with a default value of 0.05°. The specific value is calibrated based on the aerodynamic characteristics and control simulation results of the corresponding UAV. The dynamic glide angle is limited to the range of 2.5° to 3.5°.

[0085] Determine the height using the following formula. Lateral offset at the location :

[0086]

[0087] The crosswind component, The altitude of the target landing point. The vertical descent rate, For example, 2.8 m / s. The lateral offset decreases with decreasing altitude and reaches zero at the touchdown point. Wind field-compensated lateral offset is part of the determination of the original planned horizontal position of the reference waypoint when generating the main landing track, and is not part of the lateral avoidance offset applied to the avoidance waypoint after a traffic conflict occurs.

[0088] The target velocity for the final glide phase decreases linearly from the cruising velocity at the glide slope inlet to the landing velocity at the target landing point. With a tailwind, the deceleration initiation position is moved forward towards the glide slope inlet by a distance equal to 0.3 times the product of the tailwind component and the total glide time; with a headwind, the deceleration initiation position is moved backward towards the target landing point by a distance equal to 0.3 times the product of the headwind component and the total glide time. After wind field correction, the glide angle for the final glide phase remains within the range of 2.5° to 3.5°, the vertical descent rate does not exceed 5 m / s, and the touchdown velocity is within the range of 1.1 to 1.3 times the stall velocity.

[0089] Based on the dynamic glide angle, lateral offset, and target velocity variation, the original planned horizontal position, altitude, and target velocity are determined for each reference waypoint in the final glide phase. Subsequently, the first reference waypoint of the final glide phase is compared with the end waypoint of the approach cruise phase: if the positions, altitudes, speeds, and headings of both satisfy the continuity condition, they are combined to form the main landing track; otherwise, the process returns to the corresponding cruise waypoint allocation or the final glide profile generation process for local adjustments. The original planned attributes of each reference waypoint output by S200 constitute a common reference for subsequent processing by S300, S400, and S600. Subsequent traffic conflict avoidance must not mistakenly treat wind field compensation lateral offset as traffic avoidance offset.

[0090] S300: Identify turning points based on the heading difference between adjacent segments in the main landing track, and smooth the local segment where the turning point is located using the corresponding Bézier curve according to the turning level to which the heading difference belongs, to obtain the smoothed main landing track.

[0091] The system iterates through waypoints in the main landing track, calculates the heading difference between adjacent segments, and identifies waypoints with a heading difference greater than 5° as turning points. For heading differences less than 15°, a second-order Bézier curve is used for smoothing small turns; for heading differences between 15° and 60°, a single-segment third-order Bézier curve is used for smoothing medium turns, ensuring heading continuity in the turning segments; for heading differences greater than 60°, two-segment third-order Bézier curves are spliced ​​together for smoothing large turns, and the curvature continuity at the splicing positions is checked. Small and medium turns meet their respective smoothing requirements, and the large turn branch further achieves G2 continuity.

[0092] For each turning point, the waypoints before, at, and after the turning point are read to determine the inbound and outbound directions, respectively, with the local turning segment used as the smoothing object. The turning angle classification results are used to select the appropriate order and number of Bézier curves, without performing the same curve replacement on non-turning segments in the main landing track. Small turning branches generate locally smoothed curves according to second-order Bézier curves.

[0093] A single segment of a third-order Bézier curve starts from the starting point First control point Second control point and the end point definition:

[0094]

[0095] The turning point is the point of entry. The two points are located on the flight segments before and after the turning point, respectively, and are 1.2 times the turning radius away from the turning point. from Extend forward by one turning radius along the heading of the entry segment; from Extend the course of the outgoing segment by a turning radius so that the tangent direction of the curve at the start and end points is consistent with the corresponding original segment.

[0096] In parameters Curve points are obtained within a range from 0 to 1, and then converted into smooth waypoints according to the interpolation interval applicable to the local turning segment. The horizontal position of the smooth waypoint is taken from the Bézier curve, and the corresponding altitude and target speed are determined according to the original planned longitudinal attributes of the replaced local segment, thereby avoiding the regeneration of the longitudinal glide profile of the final glide segment due to horizontal turn smoothing.

[0097] For sharp turns, the turning segment is divided into two parts and a splicing point is set. The first paragraph uses , , and The second paragraph uses , , and Establish two curve segments in Constraints with equal curvature are applied, and gradient descent is used to adjust... and The position of the two curves so that they are at... The curvature difference gradually decreases at a given point; the curvature difference is less than... When the time is right, output a two-segment third-order Bézier curve.

[0098] The gradient descent method uses the curvature difference on both sides of the splicing point as the control point for adjustment. Each adjustment... and Then, the two curve segments are redefined and the curvature on both sides of the splicing point is calculated; if the curvature difference has not yet reached the threshold, the adjustment continues, and once the threshold is reached, the control point adjustment of the current local turning segment is stopped.

[0099] During the smoothing process, the maximum turning angular velocity, maximum climb rate, and maximum descent rate are set based on the UAV's current payload, altitude, and atmospheric density. If the smoothed waypoint does not meet the maneuverability limits, the turning radius is increased or the transition distance is extended before regenerating the waypoint. For turning radii less than 100m, interpolation is performed at 20m intervals; for turning radii between 100m and 500m, interpolation is performed at 50m intervals; and for turning radii greater than 500m, interpolation is performed at 100m intervals. This smoothing process only modifies the local turning segments that require smoothing and does not re-plan the longitudinal glide profile of the final glide segment.

[0100] After completing a local smoothing, the smoothed curve is reconnected to the original tracks before and after the turn, and the connection position, heading, curvature, and corresponding maneuvering limits are verified. If the verification fails, only the control point, turning radius, or transition distance of the current turn segment is adjusted; the entire main landing track is not regenerated. After all turn segments have been locally smoothed, the smoothed main landing track is output for energy verification by the S400.

[0101] S400: Based on the airspeed and altitude of each waypoint in the main landing track, determine the kinetic energy, potential energy, and mechanical energy state of each waypoint, and perform energy verification and correction on the main landing track based on the mechanical energy state.

[0102] The kinetic and potential energies of each waypoint are determined based on the airspeed and altitude in the main landing trajectory, and the mechanical energy state of the corresponding waypoint is determined based on the kinetic and potential energies. For a mass of... airspeed is Height is Waypoints:

[0103]

[0104]

[0105]

[0106] Following the waypoint sequence in the main landing track, an energy state including waypoint identifier, airspeed, altitude, kinetic energy, potential energy, and mechanical energy is generated for each waypoint. The target velocity at a waypoint is not directly defined as airspeed in this step; mechanical energy calculations use the waypoint airspeed obtained from the corresponding flight state. After completing the waypoint-by-waypoint energy calculations, mechanical energy changes and energy change rate checks are performed on adjacent waypoints in the normal final glide phase, and a grounding energy check is performed on the target landing point.

[0107] During the normal final descent, the mechanical energy state of adjacent waypoints decreases along the direction of ground contact, i.e. This decreasing relationship is not used for cruise segments where climb is permitted, nor for go-around tracks where mechanical energy increases with climb and acceleration.

[0108] The rate of energy change between adjacent waypoints is:

[0109]

[0110] Compare the absolute values ​​of the differences in energy change rates between consecutive waypoints, ensuring that this absolute value does not exceed the upper limit of the allowable difference in adjacent energy change rates set according to the aerodynamic characteristics of the UAV. , The unit is energy / time. The kinetic energy at the target landing point is expressed as... Based on this, the allowable deviation is, for example, ±5%; potential energy is... Based on this, the permissible deviation of the target landing point altitude is, for example, ±2m.

[0111] Based on the verification results, energy mismatches are categorized into kinetic energy deviation, potential energy deviation, total mechanical energy deviation, or adjacent energy change rate deviation. Mechanical energy decrease, rate of change smoothing, and ground energy matching are used to check different aspects of the track status; if any one of these is not met, the track will not be directly used as the primary landing track that passes the energy verification. The allowable upper limit is stated above. Based on the aerodynamic characteristics of the drone.

[0112] When kinetic energy is high but potential energy is normal, extend the deceleration phase or start deceleration earlier; when potential energy is high but kinetic energy is normal, adjust the descent distribution within the allowable range of the dynamic glide angle; when the overall total energy is high, add a level flight deceleration phase during the approach cruise phase; when the total energy is low, reduce the glide angle and extend the glide distance within the allowable range of the glide angle and maneuver constraints. After each adjustment, recalculate the kinetic energy, potential energy, mechanical energy, and touchdown energy of the affected waypoints. Increase deceleration when headwinds increase airspeed and kinetic energy, and delay deceleration when tailwinds decrease airspeed and kinetic energy, ensuring that the touchdown kinetic energy is within the target range.

[0113] After local correction, the mechanical energy and rate of change of energy of the affected waypoints and their adjacent waypoints are recalculated, and the corresponding verification is performed again. The correction process uses the S200's glide slope angle, rate of descent, touchdown speed, and wind field compensation constraints, and does not exceed the predetermined maneuver boundaries due to energy correction. After completing the energy verification, the position, airspeed, altitude, and mechanical energy state at the glide slope inlet decision point, glide slope midpoint decision point, and decision altitude decision point are recorded as inputs for the S500 to generate the corresponding go-around trajectory.

[0114] S500, at least one go-around decision point is set on the main landing trajectory, a corresponding go-around trajectory is pre-generated for each go-around decision point, and the association between the go-around decision point and the corresponding go-around trajectory is established.

[0115] In this embodiment, there are three decision points for the go-around: the glide path entrance decision point located at the entrance of the final glide segment, the glide midpoint decision point located in the middle of the final glide segment, and the decision altitude decision point located at a preset decision altitude within the final glide segment. The glide path entrance decision point, the glide midpoint decision point, and the decision altitude decision point correspond to the entrance energy verification point, the mid-section energy verification point, and the pre-grounding energy verification point, respectively.

[0116] The decision point at the glide slope inlet is, for example, approximately 150m high, the decision point at the mid-glide slope is, for example, approximately 80m high, and the decision height decision point is located at a preset decision height approximately 30m above the ground. These heights are set according to the UAV model, target landing point, and landing specifications.

[0117] The three go-around decision points are used both to perform complete safety checks at their respective locations and to determine the initial state of the pre-generated go-around trajectory. For each go-around decision point, the following steps are executed sequentially: initial state reading, initial climb segment generation, turn exit segment generation, waiting line segment generation, mirror constraint verification, safety constraint verification, and associated storage. The trajectory verification result of one go-around decision point does not replace the trajectory generation and verification of other go-around decision points.

[0118] For each of the aforementioned go-around decision points, the mechanical energy state at the go-around decision point is determined based on the airspeed and altitude of the main landing trajectory at that go-around decision point. The position, airspeed, and mechanical energy state of the main landing trajectory at the go-around decision point are obtained, and the position, airspeed, and mechanical energy state are used as the initial reference for the pre-generated go-around trajectory corresponding to that go-around decision point.

[0119] The initial state of the go-around trajectory corresponds to the position, airspeed, and mechanical energy state of the main landing trajectory at the go-around decision point. This ensures that the pre-generated go-around trajectory is not an arbitrarily generated backup curve independent of the main landing trajectory, but rather starts from the actual planned state of the main landing trajectory at that node. The correlation between the go-around decision point and its pre-generated go-around trajectory is established after the trajectory verification is successful, providing a trajectory corresponding to that node for subsequent go-around triggers.

[0120] The pre-generated go-around trajectory for each of the aforementioned go-around decision points includes an initial climb segment, a turning exit segment, and a waiting line segment connected in sequence; the initial climb segment is used to climb from the corresponding go-around decision point, the turning exit segment is used to turn to the waiting line direction after climbing, and the waiting line segment is used to keep the UAV in waiting flight.

[0121] The initial climb phase employs a continuous two-stage climb process. Stage A is an instantaneous transition phase, starting from the go-around decision point, with the initial position, airspeed, and mechanical energy state consistent with the main landing track at that point. Starting from the current descent state at that point, while maintaining the current heading, the vertical velocity continuously passes through zero within Stage A, transitioning to a climb rate in the opposite direction with an amplitude matching the original descent rate of the corresponding main landing track. Altitude changes continuously during this process. Stage B is a sustained climb phase, where the climb rate smoothly increases from the climb rate at the end of Stage A to the optimal climb rate for the UAV, and then maintains this optimal climb rate to a terrain-safe altitude. The optimal climb rate is, for example, 8 m / s, with its specific value determined based on the UAV model.

[0122] The go-around acceleration profile mirrors the deceleration profile of the corresponding main landing trajectory. This means the deceleration change of the main landing trajectory before the go-around switch serves as a reverse acceleration reference after the switch, ensuring consistent energy change rates during the switch and transition phases and preventing abrupt changes. This mirror relationship is used for initial state switching and smooth transition; after entering the sustained climb phase, steady-state climb is constrained by the optimal climb rate. This mirror relationship does not replace the constraints on UAV acceleration, airspeed, climb rate, and mechanical energy changes.

[0123] After climbing to a terrain-safe altitude, the UAV enters the turning exit segment. The terrain-safe altitude is set at the local maximum terrain elevation plus 100m. The turning exit segment uses a third-order Bézier curve to turn towards the waiting route, and the turning angular velocity and curvature are checked according to the same maneuver constraints as the smoothing process of the main landing trajectory. The waiting route segment is a rectangular waiting route consisting of two parallel straight sections and two circular arc turning sections, with the waiting altitude set at the local maximum terrain elevation plus 300m. The UAV maintains a preset speed and altitude in the waiting route segment in order to re-enter the landing process.

[0124] The following checks are performed: Phase A: Vertical velocity continuously passes zero from the current descent state and transitions to the opposite direction; Climb rate relationship where amplitude matches the original descent rate of the corresponding main landing trajectory; Phase B: Smooth increase in climb rate towards the optimal climb rate; Mirror relationship between the go-around acceleration profile and the main landing trajectory deceleration profile; Energy change rate during switching and transition phases; Go-around turn constraints; Terrain safety margin; Maneuver limits; and Airspace restrictions. If the corresponding check conditions are not met, the corresponding climb or acceleration profile, turn radius, or trajectory position is adjusted and regenerated. After the go-around trajectory at the three go-around decision points passes the check, it is associated with and stored at the corresponding go-around decision point.

[0125] The associated storage maintains a one-to-one correspondence between the go-around decision point and the corresponding pre-generated go-around trajectory. After processing the three fixed nodes, S500 outputs the energy status of each go-around decision point, the main landing trajectory at each node, the corresponding pre-generated go-around trajectory and its associated relationship, for S600 to call when a go-around is triggered at a fixed node.

[0126] When the go-around conditions are met at the non-go-around decision point of the final glide phase, the real-time generated emergency go-around trajectory also includes the initial climb phase, the turning exit phase, and the waiting line phase connected in sequence. It adopts the two-stage climb process, mirror acceleration relationship, turning constraints, terrain safety margin, maneuver limits, airspace restrictions, and the same verification conditions described in this step. The difference between it and the pre-generated go-around trajectory is that it uses the current position, current airspeed, and current mechanical energy state of the UAV at the time of triggering the go-around as the initial reference and generates it online when the go-around is triggered.

[0127] Online branch generation does not wait for the UAV to reach the next fixed go-around decision point, nor does it forcibly map non-fixed positions to a fixed go-around decision point. The current position, current airspeed, and current mechanical energy state obtained when a go-around is triggered directly replace the pre-stored initial state of the fixed node. Subsequent branches are still generated and verified in the order of Phase A, Phase B, turning exit segment, and waiting line segment.

[0128] S600: Obtain the predicted trajectory of the target aircraft, perform conflict prediction based on the main landing trajectory and the predicted trajectory, and determine candidate trajectories to be executed based on the conflict prediction results from the main landing trajectory, the lateral avoidance trajectory generated for the final glide phase, the pre-generated go-around trajectory associated with the go-around decision point, and the emergency go-around trajectory generated in real time at non-go-around decision points.

[0129] The predicted trajectory of the target aircraft is obtained, and a conflict prediction is performed based on the main landing trajectory and the predicted trajectory to determine whether a traffic conflict occurs in the final glide phase and the degree of the traffic conflict.

[0130] Let the UAV be object O and the target aircraft be object I. The position vectors of object O and object I in the same three-dimensional coordinate system are as follows:

[0131]

[0132] The velocity vectors are as follows:

[0133]

[0134] Determine the relative position vector and the relative velocity vector:

[0135]

[0136] when At that time, the time to reach the closest point CPA is:

[0137]

[0138] when At that time, the minimum prediction interval distance is:

[0139]

[0140] The Conflict Time (TTC) is the time from the current moment to the CPA, i.e. .when At that time, both sides are relatively stationary, let The spatial interval between the two sides remains the current real-time spatial distance. At that time, both sides are in a divergent state, which is not considered a risk of conflict, and the current interval is represented by the current real-time spatial distance.

[0141] The distance risk level and time risk level are determined separately, in descending order of risk. For the distance risk level, when... It was determined to be severe at that time; the severe distance condition was not met and It was determined to be moderate; it did not meet either the severe distance condition or the moderate distance condition. The risk level was initially determined to be mild. The boundary value of 800m was classified as moderate distance risk based on the above-mentioned orderly assessment. Regarding the time risk level, when... The time condition was determined to be severe; the time condition for severe severity was not met and The time frame was determined to be moderate; it did not meet either the severe or moderate time frame criteria and It was determined to be mild.

[0142] The traffic conflict is determined to be a minor, moderate, or severe conflict based on the higher of the distance risk level and the time risk level. When the traffic conflict occurs during the final glide phase, the minor and moderate conflicts satisfy the preset lateral avoidance conditions, and the severe conflict satisfies the preset go-around conditions.

[0143] When it is determined that no traffic conflict has occurred based on the conflict prediction, the main landing trajectory is identified as a candidate trajectory to be executed, and the candidate trajectory to be executed is provided to the S700 for follow-up safety verification.

[0144] Based on the location corresponding to the CPA and the status of both parties upon reaching that location, the segment affected by the traffic conflict is identified in the main landing track, and it is determined whether this segment belongs to the approach cruise segment or the final glide segment. The conflict segment is used to select the reference waypoints required to generate transition and avoidance offsets.

[0145] When a traffic conflict occurs during the approach cruise phase, the corresponding cruise phase avoidance method is adopted. For minor conflicts, a default rightward offset of 300m is adopted. Three waypoints are selected before and after the conflict phase, gradually increasing the offset from zero to 300m and then gradually decreasing it back to zero, while maintaining a constant cruise speed. The aircraft returns to its original route at least two waypoints before entering the final glide slope. For moderate conflicts, a rightward offset of 500m is adopted. Five transition waypoints are set before and after the conflict phase. The target speed is uniformly reduced by 10% during the conflict phase, and the aircraft returns to its original position 500m before the glide slope entrance, ensuring that the entrance position deviation does not exceed 50m and the heading deviation does not exceed 5°. For severe conflicts, an S-shaped bypass trajectory with a maximum lateral offset of 800m is generated. The turning position is smoothed using a third-order Bézier curve. The speed is reduced by 15% before the conflict, and after the bypass, the aircraft returns to its original route and speed via a 500m transition section.

[0146] When generating the cruise segment avoidance track, waypoints for entering, maintaining, and exiting the avoidance phase are determined using the original cruise navigation points before and after the conflict segment as references. For light and moderate branches, the lateral offset gradually changes according to the corresponding number of waypoints; for severe branches, a continuous S-shaped horizontal track is formed according to a third-order Bézier curve. After completing the cruise avoidance, the predicted interval between the UAV and the target aircraft, the cruise segment maneuver restrictions, and the righting status before entering the final glide phase are rechecked; if the check passes, the main landing track after completing the cruise avoidance is identified as a candidate track to be executed and provided to the S700. The cruise segment speed adjustment only applies to the corresponding cruise avoidance branch and not to the restricted lateral avoidance of the final glide phase.

[0147] When a traffic conflict occurs in the final glide phase, and the degree of conflict meets the preset lateral avoidance conditions, corresponding avoidance waypoints are generated based on at least some reference waypoints in the conflict segment corresponding to the traffic conflict in the final glide phase. Each avoidance waypoint changes only its horizontal position relative to the corresponding reference waypoint, while maintaining the original planned altitude and target speed of the corresponding reference waypoint, so as to form a lateral avoidance path that deviates from the main landing path and then converges back to the main landing path.

[0148] The mild and moderate conflicts correspond to different maximum lateral offsets, with the maximum lateral offset corresponding to the moderate conflict being greater than that corresponding to the mild conflict. When generating the lateral avoidance trajectory, if the traffic conflict is a mild or moderate conflict, the target lateral offset is determined based on the degree of the traffic conflict, and the target lateral offset is not greater than the maximum lateral offset corresponding to the degree of conflict. During the process of entering the conflict segment, the lateral offset of the avoidance waypoint relative to the corresponding reference waypoint is gradually increased to the target lateral offset, and after avoiding the traffic conflict, the lateral offset is gradually reduced to zero, so that the lateral avoidance trajectory converges back to the main landing trajectory.

[0149] Specifically, the original planned horizontal position of each reference waypoint is used as the zero offset reference, and the positive and negative directions of the lateral offset are determined according to the avoidance direction. Before entering the conflict segment, the lateral offset is gradually increased from zero to the target lateral offset according to the relative position of the selected transition waypoint in the waypoint sequence; after crossing the conflict area, the lateral offset is gradually reduced from the target lateral offset to zero, while maintaining the original planned altitude, target speed, and original glide slope parameters of the corresponding reference waypoint.

[0150] For minor conflicts, the maximum lateral offset is, for example, 150m. Four final glide paths are selected before and after the conflict segment, allowing the lateral offset to gradually increase from zero to the target lateral offset and then gradually decrease, reaching zero at the target landing point. For moderate conflicts, the maximum lateral offset is, for example, 300m. The offset begins three waypoints before the conflict point and is gradually recovered using six waypoints after the conflict ends. The resulting avoidance waypoints inherit the original planned altitude and target speed of the corresponding reference waypoints. Lateral avoidance does not regenerate the longitudinal glide profile, nor does it re-plan the original glide angle parameters of that longitudinal glide profile. Upon reaching the go-around decision point, the current mechanical energy state is determined based on the airspeed and altitude at that decision point, and the connection between the current mechanical energy state and the initial reference of the associated go-around trajectory is verified. If the offset exceeds the runway lateral safety range after a moderate conflict, the lateral avoidance trajectory is discontinued, and a go-around process is initiated.

[0151] The default avoidance direction is to the right, and the terrain, obstacles, and airspace restrictions on the right are checked; if the right is not feasible, the left is used instead. During the approach cruise phase, if neither side is feasible, a deceleration and waiting method is used to avoid conflict time. During the final glide phase, if neither side is feasible, instead of changing the target speed to wait, the current lateral avoidance trajectory is determined to be insufficient for a safe landing and a go-around is initiated.

[0152] The directional feasibility check is performed before the final candidate lateral avoidance path is formed. First, the terrain elevation, obstacle space range, and available airspace are checked based on the offset position of the right candidate; if any corresponding constraint is not met, the right candidate path is not used and a left candidate path is generated instead. If neither the left nor right candidate path can meet the constraints in the final glide phase, the safe exit relationship determined by the current claim is used to proceed with the go-around process.

[0153] Determine whether the lateral avoidance trajectory meets the preset safe landing conditions, and if it does, determine the lateral avoidance trajectory as a candidate trajectory to be executed, and provide it to S700 for follow-up safety verification.

[0154] The determination of whether the lateral avoidance trajectory meets the preset safe landing conditions includes: verifying whether the predicted minimum clearance distance between the lateral avoidance trajectory and the target aircraft meets the safe clearance requirement, whether the lateral avoidance trajectory meets the turning limit of the UAV, and whether the lateral avoidance trajectory can reconverge to the main landing trajectory; when at least one of the verifications does not meet the corresponding requirements, adjusting the target lateral offset or the transition segment length of the lateral avoidance trajectory, regenerating the lateral avoidance trajectory, and re-determining whether the regenerated lateral avoidance trajectory meets the preset safe landing conditions.

[0155] After each adjustment, the horizontal position of the avoidance waypoint is redefined, the CPA and TTC between the lateral avoidance track and the target aircraft are recalculated, and the turning limits and convergence relationship are checked again. This adjustment process is used to distinguish between "a certain check item has not been met" and "the lateral avoidance track ultimately cannot meet the conditions for safe continued landing": in the former case, the target lateral offset or transition segment length adjustment is performed first, and in the latter case, the go-around exit of claim 1 is accepted.

[0156] For collision avoidance maneuvers occurring during the approach cruise phase, the deviation from the glide slope inlet position is verified to be no more than 50m, the heading deviation no more than 5°, and the speed deviation no more than 2m / s; if these conditions are not met, the transition section is extended and the calculation is repeated. For collision avoidance maneuvers occurring during the final glide phase, the remaining lateral avoidance trajectory is verified to avoid traffic conflicts, meet turning restrictions, and converge back to the main landing trajectory at the target landing point.

[0157] When the traffic conflict occurs in the final glide phase and the conflict level meets the preset go-around conditions, or when the conflict level meets the preset lateral avoidance conditions but the lateral avoidance trajectory does not meet the preset safe landing conditions, a go-around is triggered and the landing along the final glide phase is stopped.

[0158] When the UAV is located at any of the go-around decision points when a go-around is triggered, the go-around trajectory associated with that decision point is invoked, and the invoked go-around trajectory is identified as a candidate trajectory to be executed. This invocation branch uses pre-generated go-around trajectories that have been verified and associated with the go-around decision points and stored in S500.

[0159] When the UAV is not at any of the go-around decision points when a go-around is triggered, an emergency go-around trajectory is generated in real time using the UAV's current position, current airspeed, and current mechanical energy state at the time of the go-around trigger as the initial reference. This emergency go-around trajectory is then designated as a candidate trajectory to be executed. The emergency go-around trajectory adopts the same initial climb segment, turning exit segment, and waiting line segment structure as the pre-generated go-around trajectory, as well as the same two-stage climb, mirror acceleration, maneuvering, terrain safety, airspace restrictions, and verification constraints.

[0160] Therefore, the S600 generates candidate flight paths based on conflict prediction and corresponding processing results: when no traffic conflict occurs, the original main landing flight path is used as a candidate; when a traffic conflict occurs during the approach cruise phase and is avoided and cleared, the main landing flight path after completing the cruise avoidance is used as a candidate; when a minor or moderate traffic conflict occurs during the glide phase and a safe landing is determined, the lateral avoidance flight path is used as a candidate; when the go-around conditions are met or a safe landing lateral avoidance flight path cannot be formed, the pre-generated go-around flight path or the real-time generated emergency go-around flight path is used as a candidate based on the trigger location. The S600 is only responsible for generating or obtaining the corresponding candidate flight paths and does not replace the S700's follow-up safety verification of the terrain, speed, airspace, obstacles, maneuvering, and energy conditions of the candidate flight paths.

[0161] S700, perform dynamic safety verification on the candidate flight path to be executed; when the candidate flight path to be executed passes the dynamic safety verification, output the candidate flight path to be executed as the flight path to be executed by the UAV; when the candidate flight path to be executed fails the dynamic safety verification, return to the flight path generation or correction step corresponding to the unmet safety constraint for adjustment.

[0162] The system receives candidate flight paths determined by S600 and performs follow-up safety checks on these paths. After smoothing a turn, the system checks curvature, turn rate, climb rate, and descent rate. If these are not met, the control points, turn radius, or transition distance are adjusted and the path is regenerated. After lateral avoidance, the system checks the predicted minimum separation distance, turn limits, and convergence relationship with the main landing path. If these are not met, the target lateral offset or transition segment length is adjusted. For the normal landing main path, the system checks mechanical energy change, energy change rate, and ground contact energy. If these are not met, the corresponding corrections in S400 are performed. For the go-around path, the system checks the climb rate, turn limits, terrain safety altitude, and airspace restrictions.

[0163] The follow-up verification returns to local adjustments according to the steps that generated the corresponding track results: S300 is returned when the turn smoothing result does not meet the maneuvering conditions; S400 is returned when the energy state of the normal main landing track does not meet the corresponding conditions; S500 is returned when the pre-generated go-around track or the online-generated emergency go-around track does not meet the go-around constraints; and S600 is returned when the traffic separation, turn restrictions, or convergence relationship of the lateral avoidance track does not meet the requirements. After returning to the corresponding step, only the segments or parameters related to the unmet constraints are adjusted, and the verification of this type of track is performed again. The entire planning process is not immediately determined to have failed because a certain local verification item fails to meet the requirements for the first time.

[0164] Finally, a comprehensive verification of the candidate flight paths is performed. For non-contact ground segments such as the approach cruise phase, the initial climb phase of the go-around, the turn exit phase, and the holding line phase, it is confirmed that the waypoints meet the corresponding terrain safety margins. For the final glide phase, the waypoint altitudes are verified according to the preset longitudinal glide profile, target landing point altitude, and touchdown conditions; touchdown waypoints are not required to be more than 100m above the terrain. It is also confirmed that the target speed is between the stall speed and the maximum permissible speed, the flight path does not enter restricted airspace, and maintains a safe distance from obstacles. Based on any unmet constraints, the corresponding generation or correction steps are returned for adjustment and re-verification. When the candidate flight path passes the servo safety verification, the candidate flight path is output as the flight path to be executed to the flight control system.

[0165] The comprehensive verification uses safety conditions corresponding to the functions of each flight segment, avoiding the mechanical application of terrain safety margins from non-contact ground segments to the final touchdown waypoint. When outputting the trajectory to be executed, the corresponding waypoint position, altitude, target speed, heading, and verification pass status are maintained, enabling the flight control system to execute in the waypoint sequence.

[0166] Example 2:

[0167] This embodiment provides a low-altitude unmanned aerial vehicle (UAV) landing trajectory planning system that executes the method described in Embodiment 1. The system is implemented by a processing device capable of executing the steps of the method described in Embodiment 1, including a data acquisition module, a feasibility prediction module, a main landing trajectory generation module, a turn smoothing module, a conflict prediction and avoidance module, an energy management module, a go-around trajectory generation module, and a verification output module. Each module represents a program function divided according to corresponding data dependencies.

[0168] The data acquisition module acquires the UAV's flight status information, target landing point information, wind field data, terrain data, obstacle and airspace restriction data, and the predicted trajectory of the target aircraft, and provides the corresponding data to subsequent modules. The flight status information includes the current position, current airspeed, and current altitude at the time of go-around triggering. The feasibility prediction module performs endurance, terrain, maneuver, and weather predictions based on the flight status, target landing point, remaining available energy, terrain, and weather restrictions.

[0169] Specifically, the data acquisition module provides the current location, target landing point, and remaining fuel information to the feasibility prediction module; it provides flight status, target landing point, final glide direction, and wind field data to the main landing trajectory generation module; it provides terrain, obstacle, and airspace restriction data to the corresponding trajectory generation and verification modules; and it provides the target aircraft's predicted trajectory to the conflict prediction and avoidance module. The feasibility prediction module generates endurance, terrain, maneuver, and weather prediction results respectively; if any prediction fails, subsequent processing dependent on the main landing trajectory is not initiated; if all predictions pass, the results allowing the generation of the main landing trajectory are provided to the main landing trajectory generation module.

[0170] The main landing trajectory generation module generates the main landing trajectory, which includes the approach cruise phase and the final glide phase. It corrects the longitudinal glide profile and deceleration start position based on the longitudinal wind component and generates a wind field to compensate for lateral offset based on the lateral wind component. The turn smoothing module identifies the turning points in the main landing trajectory and selects a second-order Bézier curve, a single-segment third-order Bézier curve, or a double-segment third-order Bézier curve for local smoothing based on the turning angle.

[0171] The main landing trajectory generation module outputs the original planned horizontal position, altitude, target speed, and heading of each reference waypoint, and provides the main landing trajectory to the turn smoothing module, energy management module, and conflict prediction and avoidance module. The turn smoothing module only replaces the local turn segments that need to be smoothed, outputs the smoothed main landing trajectory, and retains the longitudinal glide profile of the final glide phase reference waypoint as a reference for subsequent lateral avoidance.

[0172] The energy management module determines the kinetic energy, potential energy, and mechanical energy state based on the airspeed and altitude of each waypoint, verifies the mechanical energy change, energy change rate, and grounding energy of the normal final glide path, and provides the position, airspeed, and mechanical energy state at each go-around decision point to the go-around path generation module. When a go-around is triggered at a non-go-around decision point, the energy management module determines the current mechanical energy state based on the current airspeed and altitude at the time of triggering, and provides the current position, current airspeed, and current mechanical energy state to the go-around path generation module.

[0173] The energy management module also receives the wind field-corrected speed and altitude status and performs corresponding local corrections according to the deviation types specified in S400 when the mechanical energy verification does not meet the requirements. For fixed go-around decision points, the energy management module provides pre-calculated node states; for non-fixed locations, the energy management module provides the initial reference required for online generation based on the state at the time of triggering the go-around.

[0174] The go-around trajectory generation module pre-generates go-around trajectories including the initial climb segment, the turning exit segment, and the holding line segment for each go-around decision point, and establishes the correlation between the go-around decision point and the corresponding go-around trajectory. When a go-around is triggered at a non-go-around decision point, the module generates an emergency go-around trajectory with the same three-segment structure and generation constraints in real time, based on the go-around triggering result output by the conflict prediction and avoidance module and using the current position, current airspeed, and current mechanical energy state provided by the energy management module as the initial reference, and determines the emergency go-around trajectory as a candidate trajectory to be executed.

[0175] The conflict prediction and avoidance module determines the CPA, TTC, and conflict severity based on the main landing track and the predicted track of the target aircraft, and determines the corresponding processing branch based on the conflict prediction results. When no traffic conflict occurs, this module identifies the already generated main landing track as a candidate track to be executed. When a traffic conflict occurs during the final glide phase and lateral avoidance conditions are met, the conflict prediction and avoidance module generates avoidance waypoints that only change the horizontal position while maintaining the original planned altitude and target speed, using a reference waypoint. It then determines whether the generated lateral avoidance track meets the conditions for safe continuation of landing, and if so, identifies the lateral avoidance track as a candidate track to be executed. If the go-around conditions are met or the lateral avoidance track does not meet the conditions for safe continuation of landing, the conflict prediction and avoidance module outputs a go-around trigger result. When a go-around is triggered at a go-around decision point, this module calls the pre-generated go-around track associated with the corresponding go-around decision point and identifies the called go-around track as a candidate track to be executed.

[0176] The conflict prediction and avoidance module provides the determined candidate flight paths or go-around trigger results to subsequent modules. The go-around trigger result at a fixed go-around decision point is used to obtain a pre-generated go-around path associated with that node; the go-around trigger result at a non-fixed location, together with the current position, current airspeed, and current mechanical energy state provided by the energy management module, serves as the conditions for generating an emergency go-around path online. The conflict prediction and avoidance module does not undertake the specific generation of emergency go-around paths at non-fixed locations; this generation is performed by the go-around path generation module.

[0177] The verification output module receives candidate execution tracks provided by the conflict prediction and avoidance module or the go-around track generation module. It performs dynamic safety checks on these candidate tracks, including maneuverability, terrain, energy, conflict, and airspace checks. When a candidate track passes the dynamic safety checks, it outputs the candidate track as the execution track to the flight control system. Each module is implemented by a program function unit executed by the same processor, or distributed across multiple processing units with data communication relationships. The module division does not change the data flow and processing relationships described in Embodiment 1.

[0178] When a candidate flight path fails the corresponding verification, it is not output. Instead, the corresponding candidate flight path is re-formed and verified according to the local adjustment relationship described in Embodiment 1. After passing the follow-up safety verification, the candidate flight path is output as the flight path to be executed. Multiple modules can be executed by the same processor or by multiple processing units with data transfer relationships. The module division does not change the processing relationship described in Embodiment 1, nor the fixed node pre-generation call and the dual go-around exit generated online at non-fixed positions.

[0179] Example 3:

[0180] This embodiment provides an electronic device, including a processor and a memory. The memory stores a computer program, and when the processor executes the computer program, it implements the low-altitude UAV landing trajectory planning method described in Embodiment 1.

[0181] Example 4:

[0182] This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the low-altitude UAV landing trajectory planning method described in Embodiment 1.

[0183] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented, in whole or in part, as a computer program product, the computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0184] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for planning the landing trajectory of a low-altitude unmanned aerial vehicle (UAV), characterized in that, include: The flight status information and target landing point information of the UAV are acquired, and the main landing trajectory of the UAV is generated. The main landing trajectory includes a final glide segment, and the final glide segment includes multiple reference waypoints. Each reference waypoint has at least a horizontal position, an altitude, and a target speed. At least one go-around decision point is set on the main landing trajectory. A corresponding go-around trajectory is pre-generated for each go-around decision point, and the association between the go-around decision point and the corresponding go-around trajectory is established. Obtain the predicted trajectory of the target aircraft, perform conflict prediction based on the main landing trajectory and the predicted trajectory, and determine whether a traffic conflict occurs in the final glide phase and the degree of the traffic conflict. When it is determined that no traffic conflict has occurred based on the conflict prediction, the main landing trajectory is identified as a candidate trajectory to be executed. When the traffic conflict occurs in the final glide segment and the degree of conflict meets the preset lateral avoidance conditions, corresponding avoidance waypoints are generated based on at least some reference waypoints in the conflict segment corresponding to the traffic conflict in the final glide segment. Each avoidance waypoint only changes its horizontal position relative to the corresponding reference waypoint, while maintaining the original planned altitude and target speed of the corresponding reference waypoint, so as to form a lateral avoidance path that deviates from the main landing path and then converges back to the main landing path. Determine whether the lateral avoidance trajectory meets the preset safe landing conditions, and if it does, determine the lateral avoidance trajectory as the candidate trajectory to be executed; When the traffic conflict occurs in the final glide phase and the degree of conflict meets the preset go-around conditions, or when the degree of conflict meets the preset lateral avoidance conditions but the lateral avoidance trajectory does not meet the preset safe landing conditions, a go-around is triggered and the landing along the final glide phase is stopped. When the UAV is located at any of the go-around decision points when a go-around is triggered, the go-around trajectory associated with the go-around decision point is invoked, and the invoked go-around trajectory is determined as the candidate trajectory to be executed. When the UAV is not at any of the go-around decision points when the go-around is triggered, the current position, current airspeed and current mechanical energy state of the UAV at the time of triggering the go-around are used as the initial reference to generate an emergency go-around trajectory in real time, and the emergency go-around trajectory is determined as the candidate trajectory to be executed. Perform a follow-up safety check on the candidate flight path, and when the candidate flight path passes the follow-up safety check, output the candidate flight path as the flight path to be executed by the UAV.

2. The low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method according to claim 1, characterized in that, The generation of the main landing trajectory of the UAV includes: The wind field corresponding to the final glide path is decomposed into a longitudinal wind component along the flight direction of the final glide path and a lateral wind component perpendicular to the flight direction of the final glide path. The longitudinal glide profile of the final glide segment is corrected according to the longitudinal wind component, a wind field compensation lateral offset is formed in the final glide segment according to the lateral wind component, and the deceleration start position of the final glide segment is adjusted according to the longitudinal wind component to generate the wind field adaptive main landing trajectory. The wind field compensation lateral offset is used to determine the original planned horizontal position of each reference waypoint in the final glide segment.

3. The low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method according to claim 1, characterized in that, The step of performing conflict prediction based on the primary landing trajectory and the predicted trajectory, and determining the degree of the traffic conflict, includes: Based on the main landing trajectory and the predicted trajectory, the predicted minimum separation distance and conflict time (TTC) between the UAV and the target aircraft are determined. The distance risk level is determined based on the predicted minimum separation distance, and the time risk level is determined based on the conflict time (TTC). The traffic conflict is determined to be a minor conflict, a moderate conflict, or a severe conflict based on the higher risk level between the distance risk level and the time risk level. When the traffic conflict occurs during the final glide phase, the minor and moderate conflicts satisfy the preset lateral avoidance conditions, and the severe conflict satisfies the preset go-around conditions.

4. The low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method according to claim 3, characterized in that: The mild conflict and the moderate conflict each correspond to different maximum lateral offsets, and the maximum lateral offset corresponding to the moderate conflict is greater than the maximum lateral offset corresponding to the mild conflict. When generating the lateral avoidance trajectory, if the traffic conflict is a minor conflict or a moderate conflict, the target lateral offset is determined according to the degree of the traffic conflict, and the target lateral offset is not greater than the maximum lateral offset corresponding to the degree of conflict. During the process of entering the conflict segment, the lateral offset of the avoidance waypoint relative to the corresponding reference waypoint is gradually increased to the target lateral offset, and after avoiding the traffic conflict, the lateral offset is gradually reduced to zero, so that the lateral avoidance track converges back to the main landing track.

5. The low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method according to claim 4, characterized in that, The determination of whether the lateral avoidance trajectory meets the preset safe landing conditions includes: Verify whether the predicted minimum clearance distance between the lateral avoidance trajectory and the target aircraft meets the safety clearance requirement, whether the lateral avoidance trajectory meets the turning limit of the UAV, and whether the lateral avoidance trajectory can reconverge to the main landing trajectory; If at least one of the verifications fails to meet the corresponding requirements, adjust the target lateral offset or the transition length of the lateral avoidance trajectory, regenerate the lateral avoidance trajectory, and re-determine whether the regenerated lateral avoidance trajectory meets the preset safe landing conditions.

6. The low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method according to claim 1, characterized in that: The kinetic and potential energy of each waypoint are determined based on the airspeed and altitude of each waypoint in the main landing trajectory, and the mechanical energy state of the corresponding waypoint is determined based on the kinetic and potential energy. For each of the aforementioned go-around decision points, the mechanical energy state at the go-around decision point is determined based on the airspeed and altitude of the main landing trajectory at that go-around decision point. The position, airspeed, and mechanical energy state of the main landing trajectory at the go-around decision point are obtained, and the position, airspeed, and mechanical energy state are used as the initial reference for the pre-generated go-around trajectory corresponding to that go-around decision point.

7. The low-altitude unmanned aerial vehicle (UAV) landing trajectory planning method according to claim 6, characterized in that: The go-around decision points are three: the glide path entrance decision point at the entrance of the final glide segment, the glide midpoint decision point at the middle of the final glide segment, and the decision height decision point at the preset decision height within the final glide segment. The glide slope inlet decision point, the glide slope midpoint decision point, and the decision height decision point correspond to the inlet energy verification point, the mid-section energy verification point, and the pre-grounding energy verification point, respectively. The pre-generated go-around trajectory for each of the aforementioned go-around decision points includes an initial climb segment, a turning exit segment, and a waiting line segment connected in sequence; the initial climb segment is used to climb from the corresponding go-around decision point, the turning exit segment is used to turn to the waiting line direction after climbing, and the waiting line segment is used to keep the UAV in waiting flight.

8. A low-altitude unmanned aerial vehicle (UAV) landing trajectory planning system, characterized in that, include: The data acquisition module is configured to acquire the UAV's flight status information, target landing point information, and the predicted trajectory of the target aircraft; The main landing trajectory generation module is configured to generate the main landing trajectory of the UAV based on the flight status information and the target landing point information. The main landing trajectory includes a final glide segment, and the final glide segment includes multiple reference waypoints. Each reference waypoint has at least a horizontal position, an altitude, and a target speed. The go-around trajectory generation module is configured to set at least one go-around decision point on the main landing trajectory, pre-generate corresponding go-around trajectories for each go-around decision point, and establish the association between the go-around decision point and the corresponding go-around trajectory. The conflict prediction and avoidance module is configured to perform conflict prediction based on the main landing trajectory and the predicted trajectory, determine whether a traffic conflict occurs in the final glide segment and the degree of the traffic conflict, and when the traffic conflict occurs in the final glide segment and the degree of the conflict meets the preset lateral avoidance conditions, generate corresponding avoidance waypoints based on at least some reference waypoints in the conflict segment corresponding to the traffic conflict in the final glide segment. Each avoidance waypoint only changes its horizontal position relative to the corresponding reference waypoint, and maintains the original planned altitude and target speed of the corresponding reference waypoint, so as to form a lateral avoidance trajectory that deviates from the main landing trajectory and then converges back to the main landing trajectory. The conflict prediction and avoidance module is further configured to, when it is determined that no traffic conflict has occurred based on the conflict prediction, identify the main landing trajectory as a candidate trajectory to be executed; determine whether the lateral avoidance trajectory meets the preset safe continued landing conditions, and if it does, identify the lateral avoidance trajectory as the candidate trajectory to be executed. The conflict prediction and avoidance module is further configured to trigger a go-around and stop continuing to land along the final glide segment when the traffic conflict occurs in the final glide segment and the conflict level meets the preset go-around conditions, or when the conflict level meets the preset lateral avoidance conditions but the lateral avoidance trajectory does not meet the preset safe continuing landing conditions; when the UAV is located at any of the go-around decision points when a go-around is triggered, the go-around trajectory associated with the go-around decision point is invoked, and the invoked go-around trajectory is determined as the candidate trajectory to be executed; The go-around trajectory generation module is further configured to, when the UAV is not at any of the go-around decision points when the go-around is triggered, use the current position, current airspeed and current mechanical energy state of the UAV at the time of triggering the go-around as the initial reference, generate an emergency go-around trajectory in real time, and determine the emergency go-around trajectory as the candidate trajectory to be executed; The verification output module is configured to receive the candidate execution track, perform dynamic security verification on the candidate execution track, and output the candidate execution track as the execution track of the UAV when the candidate execution track passes the dynamic security verification.

9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the low-altitude unmanned aerial vehicle landing trajectory planning method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the low-altitude unmanned aerial vehicle landing trajectory planning method according to any one of claims 1 to 7.

Images