Aircraft pushback collision avoidance method and system based on laser point cloud

Abstract

This application relates to the field of civil aviation airport operation safety control, and in particular to an aircraft pushback collision avoidance method and system based on laser point clouds. The method includes the following steps: acquiring the laser point cloud A of the target area at the current moment; acquiring the non-ground laser point cloud A' in A; and clustering A' to obtain {A'1, A'2, ..., A'...} Q}, A' q Let {A'1, A'2, ..., A'} be the q-th cluster obtained from clustering; iterate through {A'1, A'2, ..., A'} Q If there is a category corresponding to a moving aircraft, then obtain {C1, C2, ..., C}. P If there is a moving aircraft C among them q If the direction of motion is different from the nose orientation, then predict the next moment {C1, C2, ..., C...} P The system calculates the coordinates and orientation of the nose of the aircraft and obtains the corresponding 3D bounding box; if preset conditions are met, an early warning is issued. This invention improves the safety of the aircraft rollout process.

Description

Technical Field

[0001] This invention relates to the field of civil aviation airport operation safety control, and in particular to an aircraft pushback collision avoidance method and system based on laser point clouds. Background Technology

[0002] At airports, aircraft cannot be pushed out of their parking positions on their own; they require a tow truck. The limited space between parking positions, coupled with the large size of the aircraft, makes the pushback aisle narrow, and the tow truck's view is obstructed by the aircraft, preventing it from seeing the area around the aircraft. Therefore, dedicated observers and supervisors are stationed on the tarmac to assess the risk of collisions during the pushback process.

[0003] The above-mentioned method of relying on manual judgment has the following main problems: 1) Most airports, especially busy airports, have a large number of takeoffs and landings and must operate normally under all weather conditions. Therefore, there are many situations where aircraft need to be pushed back, which requires a large number of people; 2) The method of manually judging whether a collision will occur is constrained by factors such as the ability and experience of observers and commanders, which leads to frequent collision accidents; at the same time, the large number of aircraft and vehicles moving around on the apron area also brings safety hazards to observers and commanders working on the apron. Summary of the Invention

[0004] The purpose of this invention is to provide an aircraft pushback collision avoidance method and system based on laser point clouds, so as to realize the automatic prediction of whether an aircraft will collide based on laser point clouds, and solve the safety problems of existing methods that rely on manual judgment of whether a collision will occur during the aircraft pushback process.

[0005] According to a first aspect of the present invention, an aircraft launch collision avoidance method based on laser point clouds is provided, comprising the following steps:

[0006] S100, obtain the laser point cloud A of the target area at the current time.

[0007] S200, filter A to obtain the non-terrestrial laser point cloud A' in A.

[0008] S300, cluster A' based on the distance between any two non-ground laser points in A' to obtain {A'1, A'2, ..., A'...} Q}, A' q Let q be the q-th cluster obtained by clustering, where q ranges from 1 to Q, and Q is the number of clusters obtained by clustering.

[0009] S400, iterate through {A'1, A'2, ..., A'} Q}, if {A'1,A'2,…,A' QIf a category corresponding to a moving aircraft exists in the list, then obtain {C1, C2, ..., C}. P}, and enter S500; C1 is the first moving aircraft in the target area, C2 is the second moving aircraft in the target area, C P Let P be the Pth moving aircraft in the target area, where P is the number of moving aircraft in the target area.

[0010] S500, if at the current time {C1, C2, ..., C...} P There exists a moving aircraft C in} q If the direction of movement is different from the direction of the nose, then it enters S600.

[0011] S600, based on the current time and historical time {C1, C2, ..., C... P The coordinates, orientation, and size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates of the very front of the nose and the orientation of the nose.

[0012] S700, according to {C1, C2, ..., C... P The model size information of}, the next moment {C1, C2, ..., C P The coordinates of the foremost point of the nose and the nose orientation are used to obtain the next time step {C1, C2, ..., C}. P The 3D bounding box of}.

[0013] S800, if C in the next moment q The 3D bounding box and the next time step {C1, C2, ..., C P Except for C q The distance between the 3D bounding box of any other moving aircraft and the set distance threshold, or at the next moment C q The 3D bounding box and {A'1, A'2, ..., A' Q Except for {C1, C2, ..., C} P If the distance between points in categories other than the corresponding category and the set distance threshold is less than the set distance threshold, an alert will be issued.

[0014] Compared with the prior art, the present invention has significant advantages. Through the above technical solution, the aircraft launch collision avoidance method and system based on laser point clouds provided by the present invention achieves considerable technological progress and practicality, and has broad industrial application value. It has at least the following advantages:

[0015] This invention identifies aircraft being pushed out in a target area based on laser point clouds of that area. It predicts the nose tip position and nose orientation of both the pushing-out and autonomously moving aircraft at the next moment. Based on these predicted nose tip position and orientation, a 3D bounding box is constructed for the aircraft. An early warning is issued when the distance between the pushing-out aircraft's 3D bounding box and the autonomously moving aircraft's 3D bounding box is less than a set threshold, or when the distance between the pushing-out aircraft's 3D bounding box and laser points in the non-moving aircraft category is less than a set threshold. This invention achieves automatic prediction of potential aircraft collisions. It eliminates the need for observers and control personnel on the apron during aircraft pushback to prevent collisions, thus avoiding safety hazards for these personnel and improving the safety of the aircraft pushback process. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating an aircraft ejection collision avoidance method based on laser point clouds, provided in an embodiment of the present invention;

[0018] Figure 2 A schematic diagram of the structure of an aircraft ejection collision avoidance system based on laser point clouds provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] According to a first aspect of the present invention, an aircraft pushback collision avoidance method based on laser point clouds is provided, such as... Figure 1 As shown, it includes the following steps:

[0021] S100, obtain the laser point cloud A of the target area at the current time.

[0022] According to the present invention, A = {(x1,y1,z1), (x2,y2,z2), ..., (x N ,y N ,z N)},(x n ,y n ,z n Let x be the coordinate of the nth laser point in A. n y n and z n Let x, y, and z be the x-axis coordinates, y-axis coordinates, and z-axis coordinates of the nth laser point in A, respectively. The value of n ranges from 1 to N, and N is the number of laser points in A.

[0023] It is understood that lidar can be added to airports to acquire laser point clouds of target areas. Those skilled in the art will recognize that any prior art method utilizing lidar to acquire laser point clouds falls within the protection scope of this invention.

[0024] S200, filter A to obtain the non-terrestrial laser point cloud A' in A.

[0025] According to the present invention, A' = {(x'1,y'1,z'1), (x'2,y'2,z'2), ..., (x'...} M ,y' M ,z' M )},(x' m ,y' m ,z' m Let x' be the coordinates of the m-th non-ground laser point in A'. m y' m and z' m Let x, y, and z be the x-coordinates, y-coordinates, and z-coordinates of the m-th non-ground laser point in A', respectively. The value of m ranges from 1 to M, where M is the number of non-ground laser points in A'.

[0026] Optionally, the Random Sample Consensus Algorithm (RANSAC) can be used as a robust estimation method to segment out the planar model in A and filter it out, obtaining the non-terrestrial laser point cloud A' in A. Those skilled in the art will understand that any prior art method for filtering terrestrial laser point clouds falls within the protection scope of this invention.

[0027] S300, cluster A' based on the distance between any two non-ground laser points in A' to obtain {A'1, A'2, ..., A'...} Q}, A' q Let q be the q-th cluster obtained by clustering, where q ranges from 1 to Q, and Q is the number of clusters obtained by clustering.

[0028] Optionally, A' can be clustered using a Euclidean clustering algorithm. The Euclidean clustering process is existing technology and will not be described in detail here. Those skilled in the art will understand that any clustering algorithm in the prior art falls within the protection scope of this invention.

[0029] It should be understood that {A'1, A'2, ..., A'} Q The different categories in} represent different objects, such as aircraft, vehicles, or people.

[0030] S400, iterate through {A'1, A'2, ..., A'} Q}, if {A'1,A'2,…,A' Q If a category corresponding to a moving aircraft exists in the list, then obtain {C1, C2, ..., C}. P}, and enter S500; C1 is the first moving aircraft in the target area, C2 is the second moving aircraft in the target area, C P Let P be the Pth moving aircraft in the target area, where P is the number of moving aircraft in the target area.

[0031] It should be understood that if we iterate through {A'1, A'2, ..., A'...} Q}After completion, {A'1,A'2,…,A' Q If no category for a moving aircraft is found in the list, the following steps will not be performed.

[0032] Optionally, A'1 can be matched with a pre-built model to determine whether it belongs to the aircraft category. As a specific implementation, the SHOT feature descriptor of the aircraft's point cloud model M can be pre-built, the SHOT feature descriptor of A'1 can be calculated, and the corresponding point pairs between the feature descriptions of A'1 and M can be initially estimated. The Hough voting method can be used to search for the instance I corresponding to model M in A'1. If no instance is found, A'1 is considered not to belong to the aircraft category; if an instance is found, A'1 is considered to belong to the aircraft category. The transformation matrix of instance I relative to model M can be obtained, and the initial position and nose orientation of the foremost point of the nose in instance I can be calculated using the initial position and nose orientation of the nose's foremost point in model M. The judgment of {A'2, ..., A'...} is then performed. Q The method for determining whether any category in} corresponds to the category of an aircraft can be found in the method for determining whether A'1 corresponds to the category of an aircraft.

[0033] Optionally, if A'1 is the category corresponding to an aircraft, then the methods for determining whether the aircraft C'1 corresponding to A'1 is in motion include:

[0034] S410, obtain the current time coordinate (x) of the foremost point of the nose of C'1. T ,y T ) and the historical coordinates (x) from the previous N times T-1 ,y T-1 ), (x T-2 ,y T-2 ), ..., (x T-N ,yT-N ), where x T x T-1 x T-2 and x T-N The x-axis coordinates and y-axis coordinates of the nose of the aircraft at the current time T, the first historical time T-1, the second historical time T-2, and the Nth historical time TN are respectively the x-axis coordinates and y-axis coordinates of the nose of the aircraft at the current time T, the first historical time T-1, the second historical time T-2, and the Nth historical time TN. T y T-1 y T-2 and y T-N The coordinates of the foremost point of the nose of the aircraft are C'1 at the current time T, the first historical time T-1, the second historical time T-2, and the Nth historical time TN, respectively.

[0035] It should be noted that the flight path database often does not store complete historical laser point cloud data for each aircraft (the data volume is too large), but rather the coordinates of the nose's foremost point, nose direction, and speed information at a given historical moment. Therefore, this invention determines whether an aircraft is moving based on the coordinates of the nose's foremost point at the current and previous moments. Since the nose height remains constant when an aircraft is moving on the ground, i.e., the corresponding z-axis coordinate remains constant, this invention only considers the x-axis and y-axis coordinates.

[0036] S420, if for n = 0, 1, ..., N-1, all satisfy... Then determine that C'1 is not moving; otherwise, determine that C'1 is moving; x T-0 =x T y T-0 =y T δ is a preset distance threshold.

[0037] According to the present invention, if {A'2, ..., A' Q If a certain category in the data corresponds to an aircraft, then the method for determining whether the aircraft corresponding to that category is in motion can refer to the method described above for determining whether C'1 is in motion. Those skilled in the art will understand that any existing method for identifying a specific object in a laser point cloud and for determining whether an object is in motion falls within the protection scope of this invention.

[0038] S500, if at the current time {C1, C2, ..., C...} P There exists a moving aircraft C in} q If the direction of movement is different from the direction of the nose, then it enters S600.

[0039] According to the present invention, if the direction of motion of an aircraft is different from the nose orientation, the aircraft is an aircraft being pushed back; if the direction of motion of an aircraft is the same as the nose orientation, the aircraft is an autonomously moving aircraft. The purpose of the present invention is to solve the collision problem during aircraft pushback. Therefore, the present invention determines the current time {C1, C2, ..., C...} P There exists a moving aircraft C in} q Only then can the subsequent steps of trajectory prediction and pre-collision warning be carried out.

[0040] Optionally, the method for obtaining the nose orientation includes: when determining whether a certain category corresponds to an aircraft category, upon identifying instance I of model M, a rotation and translation matrix of instance I relative to model M is obtained. Based on the known nose orientation in model M, the nose orientation in instance I can be calculated. Those skilled in the art will understand that any prior art method for determining the nose orientation based on the point cloud of an aircraft falls within the protection scope of this invention.

[0041] S600, based on the current time and historical time {C1, C2, ..., C... P The coordinates, orientation, and size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates of the very front of the nose and the orientation of the nose.

[0042] Those skilled in the art will understand that any prior art method for predicting the trajectory of an aircraft during pushback falls within the protection scope of this invention. As one existing embodiment, the above method, based on the current time and historical times {C1, C2, ..., C...},... P The coordinates, orientation, and size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates and orientation of the nose of the aircraft include:

[0043] For {C1, C2, ..., C...} P Except for C q For any moving aircraft other than those mentioned above, and for any autonomously moving aircraft: obtain the historical coordinates of the foremost point of the nose for the previous N times from the track database, and use the Kalman filter algorithm to predict the coordinates of the foremost point of the nose and the nose orientation at the next moment based on the historical coordinates, the coordinates of the foremost point of the nose at the current moment and the nose orientation.

[0044] For C q That is, for an aircraft currently being launched, predicting the next moment C. q The method for determining the coordinates of the foremost point of the nose and the nose orientation includes the following steps:

[0045] S610, based on the current time Cq The coordinates of the foremost point of the nose, the nose orientation, and the aircraft size information are used to obtain the current position of the front wheel center point P. ng (T), save it to the track database; based on C obtained from the track database q The position of the front wheel center point in the first N times {P ng (T-1), P ng (T-2), ..., P ng (TN)}, predict the position P of the front wheel center point at the next moment. ng (T+1) and the direction of motion of the front wheel center point HDG ng (T+1); Get the current time C q Direction of movement of the center point of the front wheel HDG ng (T).

[0046] It should be understood that once the nose direction is determined, C q The relative positions of the center points of the front wheels and main wheels to the foremost point of the nose are determined by the aircraft size information. Therefore, based on the current time C q Get the coordinates of the foremost point of the nose of the aircraft at the current time C. q The coordinates of the center point of the front wheel are obtained. Based on the historical positions of the front wheel center points stored in the trajectory database, the Kalman filter algorithm can be used to calculate the current direction of motion and predict the position and direction of motion at the next moment.

[0047] S620, according to P ng (T), P ng (T+1), HDG ng (T) and HDG ng (T+1) Obtain the radius r of the circular motion made by the front wheel. ng ; Obtain the center point P between the front wheel and the main wheel based on the machine size information. mg The distance between them l wb According to r ng and l wb Get P mg radius of motion r crit and combined with HDG ng (T+1)-HDG ng (T) Select the matching trajectory model.

[0048] Those skilled in the art will understand that during the aircraft rollback process, the aircraft generally moves backward in a curved path, with a small time interval between two adjacent frames of laser point clouds, and the speed of the rollback process is also relatively slow. A motion dynamics model can be applied to predict the aircraft's trajectory. Specifically, for each short segment of motion of the front wheel, it can be considered as part of the circular motion of the front wheel's center point. Therefore, based on its position and direction of motion at two different moments, the radius of the circular motion performed by the front wheel can be calculated. One of the three existing trajectory models can then be selected based on the corresponding value.

[0049] S630, Get HDG ng The angle γ between (T) and the current nose direction of the aircraft T The initial angle of the front wheel center point at the current moment is obtained based on the matched trajectory model. and final angle And predict HDG ng The angle γ between (T+1) and the nose direction at the next moment. T+1 According to HDG ng (T+1) and γ T+1 Predict the nose direction of the aircraft at the next moment.

[0050] It should be noted that the acquisition and prediction process of the relevant parameters in S610-S630 is existing technology and will not be described in detail here.

[0051] S700, according to {C1, C2, ..., C... P The model size information of}, the next moment {C1, C2, ..., C P The coordinates of the foremost point of the nose and the nose orientation are used to obtain the next time step {C1, C2, ..., C}. P The 3D bounding box of}.

[0052] According to the present invention, {C1, C2, ..., C P The size information of} is {C1, C2, ..., C} P The length, width, and height of}, C q The size information can be obtained from a pre-built model information database, {C1, C2, ..., C...} P Except for C q The additional dimensional information can be calculated based on the difference between the maximum and minimum values ​​of each coordinate axis in the corresponding point cloud and the nose direction.

[0053] According to the present invention, if the predicted C at the next moment q The x-axis coordinate of the foremost point of the nose is x0, and the predicted C for the next moment is... q The y-axis coordinate of the foremost point of the nose is y0, and the ground height is z0. The next moment, C... qLet the angle between the nose of the aircraft and the X-axis be θ, and the length, width, and height of the aircraft be L, then at the next moment C... q The coordinates of the eight vertices of the 3D bounding box are as follows: and

[0054] S800, if C in the next moment q The 3D bounding box and the next time step {C1, C2, ..., C P Except for C q The distance between the 3D bounding box of any other moving aircraft and the set distance threshold, or at the next moment C q The 3D bounding box and {A'1, A'2, ..., A' Q Except for {C1, C2, ..., C} P If the distance between laser points in categories other than the corresponding category is less than the set distance threshold, an alert will be issued.

[0055] It should be understood that by determining the next time step {C1, C2, ..., C...} P Except for C q Are the vertices of the 3D bounding box of other moving aircraft, besides C, in the next moment? q The 3D bounding box is used to determine the next time step C. q The 3D bounding box and the next time step {C1, C2, ..., C P Except for C q The system checks whether the 3D bounding boxes of other moving aircraft intersect. If they intersect, the bounding boxes are determined to be less than a preset distance threshold; otherwise, they are determined to be not less than a preset distance threshold. This is done by calculating {A'1, A'2, ..., A'...} Q Except for {C1, C2, ..., C} P For laser points in categories other than their corresponding category, the time to the next moment C q Find the minimum distance between each face of the 3D bounding box and determine whether this minimum distance is less than a set distance threshold.

[0056] Optionally, warnings can be issued by playing alarm sounds and displaying the location of potential collisions.

[0057] This invention enables automatic prediction of whether an aircraft will collide. This invention eliminates the need to add observers and command personnel on the tarmac during aircraft pushback to prevent collisions, thereby avoiding safety hazards for these personnel and improving the safety of the aircraft pushback process.

[0058] According to a second aspect of the present invention, an aircraft launch collision avoidance system based on laser point clouds is provided, such as... Figure 2As shown, the system includes a laser point cloud processing module, an aircraft trajectory prediction module, a potential conflict detection module, and an alarm display module. The output of the point cloud data processing module is connected to the inputs of the aircraft trajectory prediction module and the potential conflict detection module. The output of the aircraft trajectory prediction module is also connected to the input of the potential conflict detection module. The output of the potential conflict detection module is connected to the input of the alarm display module.

[0059] The laser point cloud processing module is used to: acquire the laser point cloud A of the target area at the current time; filter A to obtain the non-ground laser point cloud A' in A; and cluster A' according to the distance between any two non-ground laser points to obtain {A'1, A'2, ..., A'}. Q}, A' q Let {A'1, A'2, ..., A'} be the q-th cluster obtained from the clustering, where q ranges from 1 to Q, and Q is the number of clusters obtained from the clustering; iterate through {A'1, A'2, ..., A'}. Q}, if {A'1,A'2,…,A' Q If a category corresponding to a moving aircraft exists in the list, then obtain {C1, C2, ..., C}. P}, and determine the current time {C1, C2, ..., C P Does there exist aircraft C whose direction of motion is different from the direction of its nose? q C1 is the first moving aircraft in the target area, C2 is the second moving aircraft in the target area, and C... P Let P be the Pth moving aircraft in the target area, where P is the number of moving aircraft in the target area.

[0060] The aircraft trajectory prediction module is used to: based on the current time and historical times {C1, C2, ..., C...} P The coordinates, orientation, and size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates of the very front of the nose and the orientation of the nose.

[0061] The potential conflict detection module is used to: based on {C1, C2, ..., C...} P The model size information of}, the next moment {C1, C2, ..., C P The coordinates of the foremost point of the nose and the nose orientation are used to obtain the next time step {C1, C2, ..., C}. P The 3D bounding box of}; if at the next time step C q The 3D bounding box and the next time step {C1, C2, ..., C P Except for C q The distance between the 3D bounding box of any other moving aircraft and the set distance threshold, or at the next moment C qThe 3D bounding box and {A'1, A'2, ..., A' Q Except for {C1, C2, ..., C} P If the distance between points in categories other than the corresponding category is less than a set distance threshold, then a potential conflict is detected.

[0062] The alarm display module is used to issue an early warning when a potential conflict is detected.

[0063] Preferably, the alarm display module is also used to: display laser point cloud A and highlight aircraft related to potential conflicts when an alarm is triggered.

[0064] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of the invention. It should also be understood that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

1. An aircraft pushback collision avoidance method based on laser point clouds, characterized in that, Includes the following steps: S100, Obtain the laser point cloud A of the target area at the current moment; S200, filter A to obtain the non-terrestrial laser point cloud A' in A; S300, cluster A' based on the distance between any two non-ground laser points in A' to obtain {A'1, A'2, ..., A'}. Q }, A' q Let q be the q-th cluster obtained by clustering, where q ranges from 1 to Q, and Q is the number of clusters obtained by clustering. S400, iterate through {A'1, A'2, ..., A'} Q }, if {A'1,A'2,…,A' Q If a category corresponding to a moving aircraft exists in the list, then obtain {C1, C2, ..., C}. P }, and enter S500; C1 is the first moving aircraft in the target area, C2 is the second moving aircraft in the target area, C P Let P be the Pth moving aircraft in the target area, where P is the number of moving aircraft in the target area. S500, if at the current time {C1, C2, ..., C...} P There exists a moving aircraft C in} q If the direction of movement is different from the direction of the nose, then it enters S600; S600, based on the current time and historical time {C1, C2, ..., C... P The coordinates, nose orientation, and aircraft size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates of the foremost point of the nose and the orientation of the nose; S700, according to {C1, C2, ..., C... P The model size information of}, the next moment {C1, C2, ..., C P The coordinates of the foremost point of the nose and the nose orientation are used to obtain the next time step {C1, C2, ..., C}. P The 3D bounding box of}; S800, if C in the next moment q The 3D bounding box and the next time step {C1, C2, ..., C} P Except for C q The distance between the 3D bounding box of any other moving aircraft and the set distance threshold, or at the next moment C q The 3D bounding box and {A'1, A'2, ..., A' Q Except for {C1, C2, ..., C} P If the distance between laser points in categories other than the corresponding category is less than the set distance threshold, an alert will be issued.

2. The method according to claim 1, characterized in that, In S600, the step of basing decisions on the current time and historical times {C1, C2, ..., C...} P The coordinates, nose orientation, and aircraft size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates and orientation of the nose of the aircraft. include: For {C1, C2, ..., C...} P Except for C q For any other moving aircraft: obtain the historical coordinates of the nose tip of the aircraft from the track database for the previous N times, and use the Kalman filter algorithm to predict the coordinates of the nose tip and the nose orientation at the next moment based on the historical coordinates, the current coordinates of the nose tip, and the nose orientation.

3. The method according to claim 1, characterized in that, In S600, the step of basing decisions on the current time and historical times {C1, C2, ..., C...} P The coordinates, nose orientation, and aircraft size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates and orientation of the nose of the aircraft include: S610, for C q According to the current time C q The coordinates of the foremost point of the nose, the nose orientation, and the aircraft size information are used to obtain the current position of the front wheel center point P. ng (T), save it to the track database; based on C obtained from the track database q The position of the front wheel center point in the first N times {P ng (T-1), P ng (T-2), ..., P ng (TN)}, predict the position P of the front wheel center point at the next moment. ng (T+1) and the direction of motion of the front wheel center point HDG ng (T+1); Get the current time C q Direction of movement of the center point of the front wheel HDG ng (T); S620, according to P ng (T), P ng (T+1), HDG ng (T) and HDG ng (T+1) Obtain the radius r of the circular motion made by the front wheel. ng ; Obtain the center point P between the front wheel and the main wheel based on the machine size information. mg The distance between them l wb According to r ng and l wb Get P mg radius of motion r crit and combined with HDG ng (T+1)-HDG ng (T) Select the matching trajectory model; S630, Get HDG ng The angle γ between (T) and the current nose direction of the aircraft T The initial angle of the front wheel center point at the current moment is obtained based on the matched trajectory model. and final angle And predict HDG ng The angle γ between (T+1) and the nose direction at the next moment. T+1 According to HDG ng (T+1) and γ T+1 Predict the nose direction of the aircraft at the next moment.

4. The method according to claim 1, characterized in that, In S700, C q The model size information includes C q The length L, width W, and height H of the next time step; the next time step C q The coordinates of the eight vertices of the 3D bounding box are as follows: ( (), (), (), (), (), , ( )and( ); Where x0 is C at the next time step. q The x-axis coordinate of the foremost point of the nose, y0 is the C at the next moment. q The y-coordinate of the foremost point of the nose, z0 is the ground altitude, and θ is the C at the next moment. q The angle between the nose of the aircraft and the X-axis.

5. The method according to claim 1, characterized in that, In S400, A'1 is determined to be the category corresponding to the aircraft by matching it with a pre-built model.

6. The method according to claim 1, characterized in that, In S400, if A'1 is the category corresponding to an aircraft, the methods for determining whether the aircraft C'1 corresponding to A'1 is in motion include: S410, obtain the current time coordinate (x) of the foremost point of the nose of C'1. T ,y T ) and the historical coordinates (x) from the previous N times T-1 ,y T-1 ), (x T-2 ,y T-2 ), ..., (x T-N,yT-N ), where x T x T-1 x T-2 and x T-N The x-axis coordinates and y-axis coordinates of the nose of the aircraft at the current time T, the first historical time T-1, the second historical time T-2, and the Nth historical time TN are respectively the x-axis coordinates and y-axis coordinates of the nose of the aircraft at the current time T, the first historical time T-1, the second historical time T-2, and the Nth historical time TN. T y T-1 y T-2 and y T-N The y-axis coordinates of the foremost point of the nose of the aircraft at C'1, respectively, are for the current time T, the first historical time T-1, the second historical time T-2, and the Nth historical time TN; S420, if for n=0,1,…,N-1, all satisfy… If the x value is not moving, then C'1 is determined to be not moving; otherwise, C'1 is determined to be moving. T-0 =x T y T-0 =y T , This is a preset distance threshold.

7. The method according to claim 1, characterized in that, In S300, Euclidean clustering algorithm is used to cluster A'.

8. An aircraft launch collision avoidance system based on laser point clouds, characterized in that, It includes a laser point cloud processing module, an aircraft trajectory prediction module, a potential conflict detection module, and an alarm display module. The output of the laser point cloud processing module is connected to the inputs of the aircraft trajectory prediction module and the potential conflict detection module. The output of the aircraft trajectory prediction module is also connected to the input of the potential conflict detection module. The output of the potential conflict detection module is connected to the input of the alarm display module. The laser point cloud processing module is used to: acquire the laser point cloud A of the target area at the current time; filter A to obtain the non-ground laser point cloud A' in A; and cluster A' according to the distance between any two non-ground laser points to obtain {A'1, A'2, ..., A'}. Q }, A' q Let {A'1, A'2, ..., A'} be the q-th cluster obtained from the clustering, where q ranges from 1 to Q, and Q is the number of clusters obtained from the clustering; iterate through {A'1, A'2, ..., A'}. Q }, if {A'1,A'2,…,A' Q If a category corresponding to a moving aircraft exists in the list, then obtain {C1, C2, ..., C}. P }, and determine the current time {C1, C2, ..., C P Does there exist aircraft C whose direction of motion is different from the direction of its nose? q C1 is the first moving aircraft in the target area, C2 is the second moving aircraft in the target area, and C... P Let P be the Pth moving aircraft in the target area, where P is the number of moving aircraft in the target area. The aircraft trajectory prediction module is used to: based on the current time and historical times {C1, C2, ..., C...} P The coordinates, nose orientation, and aircraft size information of the nose of the aircraft are used to predict the next moment {C1, C2, ..., C}. P The coordinates of the foremost point of the nose and the orientation of the nose; The potential conflict detection module is used to: based on {C1, C2, ..., C...} P The model size information of}, the next moment {C1, C2, ..., C P The coordinates of the foremost point of the nose and the nose orientation are used to obtain the next time step {C1, C2, ..., C}. P The 3D bounding box of}; if at the next time step C q The 3D bounding box and the next time step {C1, C2, ..., C P Except for C q The distance between the 3D bounding box of any other moving aircraft and the set distance threshold, or at the next moment C q The 3D bounding box and {A'1, A'2, ..., A' Q Except for {C1, C2, ..., C} P If the distance between points in categories other than the corresponding category is less than a set distance threshold, then a potential conflict is detected. The alarm display module is used to issue an early warning when a potential conflict is detected.

9. The system according to claim 8, characterized in that, The alarm display module is also used to: display laser point cloud A and highlight aircraft related to potential conflicts when an alarm is triggered.

Images