A vehicle beyond-line-of-sight perception method for surface-oriented aircraft-road-vehicle fusion

Abstract

This disclosure relates to a vehicle beyond-line-of-sight perception method for surface aircraft based on road-to-vehicle fusion. The method determines the surface aircraft pose based on signals detected by roadside lidar. When the vehicle-mounted lidar cannot detect the surface aircraft's position information, but the roadside lidar can, a Kalman filter is performed to fuse the vehicle's GPS positioning information and the roadside lidar-detected vehicle position information to determine the vehicle's position information. When the vehicle-mounted lidar cannot detect the surface aircraft's position information, the surface aircraft pose can be obtained from the cloud via the airport communication network, thus achieving vehicle beyond-line-of-sight perception. The Kalman filter fusion of the vehicle-mounted GPS positioning information and the roadside lidar-detected vehicle position information improves the perception accuracy of the vehicle's position guided by surface services.

Description

Technical Field

[0001] This disclosure belongs to the field of road-vehicle fusion systems, and specifically relates to a vehicle beyond-line-of-sight perception method for road-vehicle fusion of surface aircraft. Background Technology

[0002] With the continuous development of intelligent connected vehicle technology and industry, intelligent connected vehicles are also playing an important role in airport surface service guidance.

[0003] However, due to the limitations of the intelligent connected vehicle's own environmental perception system in processing surface information, it is unable to accurately perceive the precise location of surface aircraft and confirm the safe distance between the service guidance vehicle and the surface taxiing aircraft. Ensuring the safe and efficient driving of the surface service guidance vehicle is a key issue in the vehicle-road cooperative roadside perception fusion for airport surface aircraft. Summary of the Invention

[0004] This disclosure is made based on the aforementioned needs of the prior art. The technical problem to be solved by this disclosure is to provide a vehicle beyond-line-of-sight perception method for surface-oriented aircraft-road-vehicle fusion.

[0005] To address the aforementioned problems, the technical solutions provided in this disclosure include:

[0006] The system obtains the location information of surface aircraft detected by vehicle-mounted LiDAR or the location information of surface aircraft detected by roadside LiDAR from the airport communication network; determines the attitude of surface aircraft based on the signals detected by roadside LiDAR; when vehicle-mounted LiDAR detects the location information of surface aircraft, the system determines the vehicle's location information using the positioning information provided by vehicle-mounted GPS; when vehicle-mounted LiDAR cannot detect the location information of surface aircraft but roadside LiDAR can, the system performs Kalman filtering fusion based on vehicle-mounted GPS positioning information and vehicle location information detected by roadside LiDAR to determine the vehicle's location information; and displays the determined location information of surface aircraft, the attitude of surface aircraft, and the location information of the vehicle in real time on the terminal.

[0007] Preferably, the step of determining the attitude of the surface aircraft based on the surface aircraft signals detected by the roadside lidar includes: S21 obtaining the surface aircraft position information detected by the roadside lidar from the airport communication network, and determining the coordinates of the aircraft's tire points under the roadside lidar coordinates. The coordinates of the two ends of the wingspan are The coordinates of the nose are S22 sets the coordinates of the aircraft tire points in the roadside lidar coordinate system as follows: The coordinates of the wingspan at both ends are The coordinates of the nose are Converting to geocentric coordinates of the aircraft tire points is The coordinates of the wingspan at both ends are The coordinates of the nose are ;

[0008] The plane containing the multiple tire points of the aircraft described in S23 is the first plane, and the plane defined by the endpoints of the wing ends and the nose end is the second plane. The position of the aircraft is determined by the angle θ between the first plane and the second plane.

[0009] Preferably, when the vehicle-mounted lidar cannot detect the location information of the surface aircraft but the roadside lidar can, the method for determining the parameters of the Kalman filter fusion model in the Kalman filter fusion process based on the vehicle-mounted GPS positioning information and the vehicle location information detected by the roadside lidar includes:

[0010] S31 determines the state of the k-th acquisition point in the Kalman filter fusion. With measurement The relationship is determined by the formula: in, Here is the state transition matrix. To control the quantity, The matrix representing the effect of the control variables; This is the observation matrix. For process noise, For measuring noise;

[0011] S32 settings , The mean is 0, and the covariances are respectively Normal distribution, predicting state And predicting and estimating the covariance matrix Determined by the formula: Specifically, the collection points are updated to obtain the predicted state of the Kth collection point. And predicting and estimating the covariance matrix : in, This represents the Kalman filter gain.

[0012] S33 is based on the Kalman filter principle and sets initial values. and Based on the measurement at time k To obtain the state at time k. ; in, Let x be the x-coordinate of the vehicle's GPS receiver in the coordinate system at time k. Let be the ordinate in the GPS receiver coordinate system at time k. Let x be the x-coordinate in the GPS receiver coordinate system at time k-1. Let y be the ordinate in the GPS receiver coordinate system at time k-1. , Let be the rotation and translation matrix components of the roadside lidar at time k. Here are the rotation and translation matrix components of the vehicle-mounted GPS at time k; the state transition matrix. for: Measurement Set as: Where Gxk and Gyk are the x and y coordinates in the vehicle coordinate system. Observation matrix for:

[0013] The control quantity's action matrix Set to 0, covariance covariance ,in It is a constant. It is a sixth-order identity matrix.

[0014] Preferably, before determining the vehicle's location information by fusing the vehicle's GPS positioning information and the vehicle's location information detected by the roadside lidar using Kalman filtering, the method further includes: performing spatiotemporal registration of the vehicle's GPS positioning information and the vehicle's location information detected by the roadside lidar based on a chain synchronization and feature matching method.

[0015] Compared to existing technologies, this solution enables beyond-line-of-sight (BLOS) perception of surface service guidance vehicles when the vehicle-mounted LiDAR cannot detect the location information of surface aircraft. This is achieved by acquiring the aircraft's pose from the cloud via the airport communication network. The vehicle-mounted LiDAR has a detection radius of 200 meters, and the roadside LiDAR also has a detection radius of 200 meters, effectively extending the vehicle's detection range by an additional 200 meters, for a total of 400 meters. Kalman filtering is used to fuse the vehicle's GPS positioning information and the location information detected by the roadside LiDAR to determine the vehicle's position. This fusion improves the accuracy of the surface service guidance vehicle's location perception. Utilizing the airport communication network, information can be accurately, timely, and quickly shared between the aircraft cockpit, control tower, surface vehicles, airlines, and airport operations control departments via a dedicated aviation 5G network. This increases data sources and volume, improves communication stability and robustness, and ensures the safe and efficient operation of the surface service guidance vehicles. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 : This is a flowchart of a vehicle beyond-line-of-sight perception method for surface-oriented aircraft-road-vehicle fusion provided in this specific embodiment;

[0018] Figure 2 : This is a flowchart of the method for determining the attitude of surface aircraft based on surface aircraft signals detected by roadside lidar provided in this specific embodiment;

[0019] Figure 3 : A flowchart illustrating the method for determining the parameters of the Kalman filter fusion model provided in this specific embodiment. Detailed Implementation

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

[0021] In the description of the embodiments of this disclosure, it should be noted that, unless otherwise expressly specified and limited, the term "connected" should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this disclosure according to the specific circumstances.

[0022] To facilitate understanding of the embodiments of this application, the following will provide further explanation and description with reference to the accompanying drawings and specific embodiments. These embodiments do not constitute a limitation on the embodiments of this application.

[0023] With the continued rapid growth of air traffic, the efficient and safe operation of airport surface service vehicles has become particularly important to ensure the orderly takeoff and landing of all aircraft. The limitations of the intelligent connected vehicles' own environmental perception systems in processing information for specific road traffic scenarios necessitate the assistance of roadside perception technology that integrates road and vehicle systems for safer and more efficient operation. Therefore, this invention provides a vehicle beyond-line-of-sight perception method for surface aircraft using road-vehicle fusion.

[0024] Specifically, the flowchart of a vehicle beyond-line-of-sight perception method for surface aircraft-road-vehicle fusion provided in this embodiment is as follows: Figure 1 As shown, it includes the following steps:

[0025] S10 obtains the location information of surface aircraft detected by the vehicle-mounted lidar; and obtains the location information of surface aircraft detected by the roadside lidar from the airport communication network.

[0026] The road-vehicle integration system used in this solution is 5G AeroMACS2.0 (Aeronautical 5G Airport Surface Broadband Mobile Communication System), a new generation of aviation broadband communication technology that applies fifth-generation mobile communication technology (5G) to the AeroMACS civil aviation dedicated network. Aircraft cockpits, control towers, surface vehicles, airlines, and airport operations control departments can all accurately, timely, and quickly share information through the aviation 5G private network, such as high-precision digital maps of the airport, runway, taxiway, jet bridge, and parking stand occupancy status, real-time aircraft and vehicle locations, and taxiing routes published by the control tower.

[0027] Vehicle GPS information, airborne GPS information, and surface aircraft location information detected by roadside lidar are uploaded to the cloud platform database through AeroMACS2.0. Therefore, the surface aircraft location information detected by roadside lidar can be obtained from the airport communication network AeroMACS2.0.

[0028] S20 determines the attitude of surface aircraft based on signals detected by roadside lidar. The flowchart of the method is as follows: Figure 2 As shown.

[0029] S21 obtains the surface aircraft position information detected by the roadside lidar from the airport communication network, and determines the tire coordinates of the aircraft in the roadside lidar coordinate system. The coordinates of the wingspan at both ends are The coordinates of the nose are .

[0030] The point cloud of the aircraft after scanning the surface was obtained by roadside lidar. The position of the aircraft provided by AeroMACS2.0 was used as the initial position. The point cloud was matched by the NDT algorithm to obtain the map coordinates of six points: three coordinates of the aircraft tires, two coordinates of the two wingspan segments, and one coordinate of the nose.

[0031] The space occupied by the prior reference point cloud (airport map) is divided into voxels of a specified size. Within each voxel, there is a set of laser point clouds. Its mean is The covariance matrix is The probability of obtaining a laser point within this voxel is: Calculate the corresponding NDT registration score. For the point cloud to be registered, consider transforming it into the mesh of the reference point cloud using transformation T: remember ,but and Continue to find the second partial derivative of s: According to the transformation equation, the vector of the second derivative of vector q with respect to the transformation parameter p is:

[0032] The iterative calculation continues until operator T converges, at which point the best match for the current frame's point cloud information is obtained, along with the map coordinates of six points in the roadside lidar coordinate system: the aircraft tires, the wingspan ends, and the nose. Let the aircraft tire coordinates be... The coordinates of the wingspan at both ends are The nose coordinates are S22 sets the coordinates of the aircraft tire points in the roadside lidar coordinate system as follows: The coordinates of the wingspan at both ends are The coordinates of the nose are Converting to geocentric coordinates of the aircraft tire points is The coordinates of the wingspan at both ends are The coordinates of the nose are .

[0033] The coordinates of the aircraft tire point under the roadside lidar coordinates are: The coordinates of the wingspan at both ends are The nose coordinates are Since the location of the roadside lidar is fixed, this transformation step is equivalent to rotating the original coordinates and then adding a fixed translation.

[0034] Describe the scanning coordinate system The origin 0 is the laser emission point, the X-axis points in the direction of the aircraft's movement, the Y-axis is vertically upward, and the Z-axis is perpendicular to the X-axis, forming a right-handed coordinate system; Earth-centered, Earth-fixed coordinate system. The origin O is the Earth's center of mass. The Z-axis is parallel to the Earth's axis and points towards the North Pole. The X-axis points towards the intersection of the Prime Meridian and the Equator. The Y-axis is perpendicular to the XOZ plane (i.e., the intersection of 90° East longitude and the Equator), forming a right-handed coordinate system. The transformation formula is as follows: Let the coordinates of the aircraft tire point after transformation be... The coordinates of the nose are The coordinates of the wingspan at both ends are .

[0035] The plane containing the multiple tire points of the aircraft described in S23 is the first plane, the planes defined by the endpoints of the wing ends and the nose end are the second plane, and the position of the aircraft is determined by the angle θ between the first plane and the second plane. in

[0036] S30 obtains vehicle GPS location information.

[0037] When the vehicle-mounted lidar detects the location information of an aircraft on the ground, the vehicle's location information is determined using the positioning information provided by the vehicle-mounted GPS.

[0038] When the vehicle-mounted LiDAR cannot detect the location information of the surface aircraft, but the roadside LiDAR can, the Kalman filter fusion process based on the vehicle's GPS positioning information and the vehicle's location information detected by the roadside LiDAR will be performed. Before the Kalman filter fusion, a chain synchronization and feature matching method is required to perform spatiotemporal registration of the vehicle's GPS positioning information and the vehicle's location information detected by the roadside LiDAR. Chain synchronization refers to the information obtained by the roadside LiDAR and the information obtained by the vehicle's GPS being at the same point in time. Feature matching refers to the point cloud registration of the roadside LiDAR point cloud using the NDT algorithm. The flowchart of the method for determining the parameters of the Kalman filter fusion model is shown below. Figure 3 As shown: S31 determines the state of the k-th acquisition point in the Kalman filter fusion. With measurement The relationship is determined by the formula: in, Here is the state transition matrix. To control the quantity, The matrix representing the effect of the control variables; This is the observation matrix. For process noise, For measuring noise.

[0039] S32 settings , The mean is 0, and the covariances are respectively Normal distribution, predicting state And predicting and estimating the covariance matrix Determined by the formula: Specifically, the collection points are updated to obtain the predicted state of the Kth collection point. And predicting and estimating the covariance matrix : in, This represents the Kalman filter gain.

[0040] S33 is based on the Kalman filter principle and sets initial values. and Based on the measurement at time k To obtain the state at time k. .

[0041] Considering two adjacent time frames, if the distance between them is sufficiently short, the vehicle position conforms to the formula for uniformly accelerated motion, that is: Therefore, the state variables of the Kalman filter are defined as the x-coordinate and y-coordinate of the vehicle's GPS receiver at time k and (k-1) respectively. Let be the rotation and translation matrix components of the lidar at time k. Let be the rotation and translation matrix components of GPS at time k.

[0042] in, Let x be the x-coordinate of the vehicle's GPS receiver in the coordinate system at time k. Let be the ordinate in the GPS receiver coordinate system at time k. Let x be the x-coordinate in the GPS receiver coordinate system at time k-1. Let y be the ordinate in the GPS receiver coordinate system at time k-1. Let be the rotation and translation matrix components of the roadside lidar at time k. Here are the rotation and translation matrix components of the vehicle-mounted GPS at time k; the state transition matrix. for:

[0043] Measurement Set as: Where Gxk and Gyk are the x and y coordinates in the vehicle coordinate system. Observation matrix for: The control quantity's action matrix Set to 0, covariance = covariance ,in It is a constant. It is a sixth-order identity matrix.

[0044] At this point, all parameters in the Kalman filter have been determined, realizing the fusion of vehicle positioning based on onboard GPS and roadside lidar detection.

[0045] If a vehicle-mounted LiDAR can detect the location information of surface aircraft, then the vehicle's GPS positioning information is the vehicle's location information. When a surface aircraft is outside the detection range of the vehicle's LiDAR, the vehicle information is determined by fusing the vehicle's GPS positioning information and the vehicle's location information detected by the roadside LiDAR, thus achieving vehicle beyond-line-of-sight perception.

[0046] S40 displays the determined surface aircraft position information, the surface aircraft attitude, and the vehicle position information in real time on the terminal.

[0047] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only a specific embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for vehicle beyond visual range perception for scene-oriented aircraft road vehicle fusion, characterized in that, The method comprises the following steps: obtaining the airport position information of the airport detected by the vehicle-mounted laser radar from the vehicle-mounted laser radar, or obtaining the airport position information of the airport detected by the roadside laser radar from the airport communication network; determining the airport pose according to the airport signal detected by the roadside laser radar; when the vehicle-mounted laser radar detects the airport position information, determining the position information of the vehicle according to the positioning information provided by the vehicle-mounted GPS; when the vehicle-mounted laser radar cannot detect the airport position information while the roadside laser radar can detect the airport position information, performing Kalman filtering fusion based on the vehicle-mounted GPS positioning information and the vehicle position information detected based on the roadside laser radar to determine the position information of the vehicle, comprising: The chain synchronization and feature matching method is used to perform space-time registration on the vehicle GPS positioning information and the vehicle position information detected by the roadside laser radar, the chain synchronization refers to that in the time sequence, the information obtained by the roadside laser radar detection and the information obtained by the vehicle GPS are at the same time point, the feature matching refers to that the point cloud registration of the roadside laser radar point cloud is realized through the NDT algorithm; S31 determines the state of the kth acquisition point fused by the Kalman filter The relationship with the measurement is determined by the formula: Wherein, is a state transition matrix, is a control quantity, is a control quantity matrix; is an observation matrix, is a process noise, is a measurement noise; S32 set , The mean is 0, and the covariance is Normal distribution, the predicted state And the predicted estimation covariance matrix Determined by the formula: Where, update the collection points, get the Kth collection point predicted state And the predicted estimation covariance matrix : Where, Kalman filter gain; S33 setting initial value based on Kalman filtering principle and , according to the measurement of the kth moment , get the state of the kth moment ; wherein, is the horizontal coordinate of the kth moment of the vehicle GPS receiver coordinate system, is the longitudinal coordinate of the kth moment of the GPS receiver coordinate system, is the horizontal coordinate of the k-1th moment of the GPS receiver coordinate system, is the longitudinal coordinate of the k-1th moment of the GPS receiver coordinate system, , is the rotation and translation matrix component of the roadside laser radar at the kth moment, , is the rotation and translation matrix component of the vehicle-mounted GPS at the kth moment; the state transition matrix is: measurement is set as: wherein, Gxk, Gyk are the horizontal and longitudinal coordinates in the vehicle coordinate system; the observation matrix is: the action matrix of the control quantity is set to 0, the covariance = , the covariance , wherein is a constant, is a six-order unit matrix; the determined airport position information, the airport pose and the vehicle position information are displayed in real time on the terminal.

2. The vehicle over-the-horizon perception method for aircraft-road integration of claim 1, wherein, The scene aircraft pose is determined according to the scene aircraft signal detected by the roadside laser radar, comprising: S21 obtaining the scene aircraft position information detected by the roadside laser radar from the airport communication network, determining the aircraft tire point coordinates in the roadside laser radar coordinates as , the wing span both end coordinates as , and the nose coordinates as ; S22 converting the aircraft tire point coordinates in the roadside laser radar coordinates as , the wing span both end coordinates as , and the nose coordinates as into the aircraft tire point coordinates in the geocentric coordinate system as , the wing span both end coordinates as , and the nose coordinates as ; S23 the plane where the plurality of tire points of the aircraft are located is a first plane, the determined plane where the wing span both ends and the nose end point are located is a second plane, and the aircraft position is determined by including the included angle θ between the first plane and the second plane:

3. The vehicle over-the-horizon perception method for aircraft-road integration of claim 1, wherein, before the Kalman filtering fusion based on the vehicle-mounted GPS positioning information and the vehicle position information detected based on the roadside laser radar to determine the position information of the vehicle, the method further comprises: performing space-time registration on the vehicle-mounted GPS positioning information and the vehicle position information detected by the roadside laser radar based on the chain synchronization and feature matching method.

Images