WO2020228228A1 - 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法 - Google Patents

基于雷达轨迹构建机坪场面运动目标运行意图识别的方法 Download PDF

Info

Publication number
WO2020228228A1
WO2020228228A1 PCT/CN2019/110514 CN2019110514W WO2020228228A1 WO 2020228228 A1 WO2020228228 A1 WO 2020228228A1 CN 2019110514 W CN2019110514 W CN 2019110514W WO 2020228228 A1 WO2020228228 A1 WO 2020228228A1
Authority
WO
WIPO (PCT)
Prior art keywords
apron
scene
track
model
target
Prior art date
Application number
PCT/CN2019/110514
Other languages
English (en)
French (fr)
Inventor
庄青
邵明珩
张震亚
张钟灵
苏祖辉
黄琰
章昆
王钟慧
Original Assignee
南京莱斯信息技术股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 南京莱斯信息技术股份有限公司 filed Critical 南京莱斯信息技术股份有限公司
Publication of WO2020228228A1 publication Critical patent/WO2020228228A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2458Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2471Distributed queries
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • G06F18/241Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • GPHYSICS
    • G08SIGNALLING
    • G08GTRAFFIC CONTROL SYSTEMS
    • G08G5/00Traffic control systems for aircraft, e.g. air-traffic control [ATC]
    • G08G5/04Anti-collision systems
    • GPHYSICS
    • G08SIGNALLING
    • G08GTRAFFIC CONTROL SYSTEMS
    • G08G5/00Traffic control systems for aircraft, e.g. air-traffic control [ATC]
    • G08G5/06Traffic control systems for aircraft, e.g. air-traffic control [ATC] for control when on the ground
    • G08G5/065Navigation or guidance aids, e.g. for taxiing or rolling

Definitions

  • the invention belongs to the technical field of airport ramp control automation of civil aviation air traffic control (ATC), and specifically relates to a method for constructing an apron scene moving target operation intention recognition method based on a radar track.
  • ATC civil aviation air traffic control
  • A-SMGCS advanced surface activity guidance and control systems
  • A-SMGCS advanced surface activity guidance and control systems
  • the scene conflict warning cannot meet the requirements of the controller in advance warning. Therefore, avoiding aircraft intrusion on the operating runway and avoiding various operating conflicts on the taxiway mainly rely on the controller passing the scene Surveillance radar and visual observation are completed.
  • the method of the present invention combines the power and kinematics models of the aircraft or vehicle, and analyzes the speed and position of the aircraft or vehicle during the entire navigation stage of the surface activity through the massive radar trajectory data based on Hadoop to construct the scene motion Target operation intention recognition model and real-time correction using actual trajectory data.
  • the relevant verification work is completed in the actual engineering project, which lays the foundation for improving the prediction ability of surface movement target operation, so as to resolve potential conflicts in advance and protect the flight scene Safe operation.
  • the purpose of the present invention is to provide a method for recognizing the operation intention of moving targets on the apron scene based on radar trajectories, so as to solve the problem of recognizing the operation intention and movement of aircraft or vehicles by the controller in the prior art. It is difficult to distinguish between various types of aircraft or vehicles in the actual environment by manually predicting the position, and the problem of large errors in the recognition of operational intentions and prediction of motion trajectories, high deviation of the prediction results, and low data availability.
  • the method of the present invention for constructing the operation intention recognition of the moving target on the apron scene based on the radar trajectory includes the following steps:
  • Step 2) According to the apron scene site environment, construct the scene operation intention model of the site. Based on the establishment of the navigation trail sample library (referred to as the navigation trail sample library), the navigation trail sample library Correlate with the scene operation intention model, mark out the operation intention model category of the navigation track sample database, and establish the moving target operation intention recognition model of the airport apron scene (referred to as this field).
  • the navigation trail sample library referred to as the navigation trail sample library
  • the navigation trail sample library Correlate with the scene operation intention model, mark out the operation intention model category of the navigation track sample database, and establish the moving target operation intention recognition model of the airport apron scene (referred to as this field).
  • step 1) specifically includes:
  • the operational target voyage record information in step 11) includes: operational target type (aircraft or vehicle), model (model or vehicle type), task type number (flight number or task order number), marking number (aircraft tail No./Landing time or license plate number/Task release time), track point, coordinate position, passing time, passing speed.
  • the label number if the aircraft is composed of "tail number” and "departure time”, if it is an inbound flight, the landing time will be adopted, and if it is an outbound flight, the departure time will be adopted; ”And “Task release time”.
  • trajectory data and routing plan are divided into two types: aircraft and vehicles;
  • Track data (multiple pieces of data): target type (aircraft), task type number (flight number), track point, coordinate position, transit time, transit speed;
  • Routing plan single piece of data: target type (aircraft), model (model), task type number (flight number), tail number, departure and landing time, gear shift time (or gear withdrawal time);
  • the trajectory data and the routing plan are associated with the target type and the task type number. For example, the flight track data is recorded and the "passing time” is recorded.
  • the routing plan records the data between the "landing time” and the “shifting time” ; If departing from the port, take the route plan to record the data between the "withdrawal gear time” and "departure time”.
  • Track data (multiple data): target type (vehicle), task type number (task order number), track point, coordinate position, passing point time, passing point speed;
  • Routing plan single piece of data: target type (vehicle), model (model), task type number (task order number), license plate number, task release time, task end time;
  • the trajectory data and the routing plan are associated with the target type and the task type number, and the trajectory data is taken to record the "time over time” and the routing plan records the data between the "task release time” and the "task end time”.
  • the step 13) specifically includes: the radar track three-dimensional position observation data adopts the WGS-84 coordinate system, B K is the longitude of the K track point in the WGS-84 coordinate system, and L K is the WGS-84 coordinate system.
  • the latitude of the track point, H K is the height of the track point K in the WGS-84 coordinate system; the track is expressed as:
  • X K is an ECEF coordinate system x-axis value
  • Y K is ECEF coordinate system y-axis values
  • Z K ECEF coordinate system is a z-axis value
  • N e is a front vertical radius of curvature, e is the eccentricity of the earth ellipsoid
  • a is the semi-major axis of the earth's ellipsoid, that is, the radius of the earth's equator, a is 6,378,137 meters;
  • b is the semi-minor axis of the earth's ellipsoid, that is, the polar radius of the earth, and b is 635,6752.3 meters;
  • the origin is the center of mass of the earth, and the track is expressed as:
  • the acceleration a k is used to describe the acceleration and deceleration movement characteristics of the sailing track on the apron scene, the formula is as follows:
  • V K and T K are respectively the "pass point speed” and "pass point time” in the voyage record information of the surface operation target.
  • step 2) specifically includes:
  • the movement intention change of the aircraft or vehicle is basically in the intersection area, with the intersection as the center, and the nearby area is set as the movement intention recognition area;
  • the big data distributed system architecture is used to replace the traditional system architecture, which solves the problem that the traditional system architecture is difficult to operate on massive data, and obtains the operation results efficiently;
  • the recognition model obtained through the support of big data replaces manual experience, combined with various types of aircraft and vehicle kinematics models, refines the classification of recognition models, reduces the unity of recognition attributes of the intention model, and improves the accuracy of prediction;
  • the intention recognition model is continuously revised in real time based on actual data to further improve the accuracy of navigation prediction, make early plans for the next work, greatly reduce or even avoid scene conflicts, and improve scene operation safety.
  • High-precision intention recognition and navigation prediction will improve the level of traffic safety and efficiency at the same time. To a certain extent, it will also increase the surface operation flow, reduce the workload of the controller, and improve the air transport service capability.
  • Figure 1 is a road map of the airport apron scene
  • Figure 2 shows the intention of the moving target moving from the intersection R11 through the four-fork road R1;
  • Figure 3 is an example of a model diagram of the running target identifying the running intention with acceleration a k ;
  • Figure 4 shows the running target in the example with the heading angle Identify the operational intent model diagram
  • Figure 5 is a diagram of a geocentric and ground-fixed rectangular coordinate system.
  • the method of the present invention combines aircraft or vehicle dynamics and kinematics models, and uses a large data Hadoop distributed computing framework based on massive surface radar trajectory data to analyze the speed, position and other information of the aircraft or vehicle during the entire navigation phase of surface activities.
  • Establish a target feature set that combines target type, model, task and other attributes and acceleration, trajectory angle, motion stage and other motion characteristics; through offline training and online testing of target feature set and operation intention information, carry out the taxiing route of the target scene Intent analysis, constructing a recognition model for the operation intention of moving targets on the apron scene.
  • the model involves the main key factors including target type, model, mission attribute, movement intention, position, heading angle (that is, the distance between the vertical axis of the aircraft and the space shuttle and the north pole of the earth). Included angle) and acceleration recognition range, etc., and use the motion intention recognition model of the present invention to correct the trajectory prediction in real time in real scenes, and achieve very good results, which lays the foundation for the research of new and comprehensive scene aircraft and vehicle trajectory prediction.
  • the method for constructing a target operation intention recognition model based on the radar trajectory of a moving target on an apron scene of the present invention includes the following steps:
  • Step 1) Hadoop-based aircraft or vehicle operation target feature set; the specific steps are as follows:
  • each apron surface voyage record consists of the operating target type (aircraft or vehicle), model (model or model), and task type number (flight number or task order number) , Marking number (tail number/landing time or license plate number/task release time), track point, coordinate position, passing point time, passing point speed and other attributes.
  • Delete invalid records with empty attributes in the record add the serial number field, and fill in the value for this activity target in order of the past time during the scene operation.
  • Migrate the scene voyage records in nine dimensions of serial number, operation target type, model, task type number, label number, track point, coordinate position, transit time, and transit speed to the distributed database HBase of the Hadoop cluster;
  • the Map process maps the original data stored in HBase to the voyage record information related to two track points.
  • the specific information items of the intermediate data of the Map are shown in Table 2;
  • the radar track three-dimensional position observation data adopts the WGS-84 coordinate system (WGS-84 coordinate system is the global coordinate system adopted globally, and the GPS broadcast satellite
  • WGS-84 coordinate system is the global coordinate system adopted globally, and the GPS broadcast satellite
  • B K is the longitude of the K track point in the WGS-84 coordinate system
  • L K is the latitude of the K track point in the WGS-84 coordinate system
  • H K is the WGS-84 coordinate Is the height of the K track point.
  • the track is expressed as:
  • Figure 5 shows the Earth-Centered, Earth-Fixed, or ECEF for short. It is a ground-fixed coordinate system (also known as the Earth coordinate system) with the center of the earth as its origin. ECEF coordinates The system is firmly connected to the earth and rotates with the earth.
  • the origin O(0,0,0) is the center of mass of the earth
  • the z-axis is parallel to the earth's axis and points to the north pole
  • the x-axis points to the intersection of the prime meridian and the equator
  • the y-axis is perpendicular to the xOz plane (that is, the intersection of 90 degrees east longitude and the equator).
  • Right-handed coordinate system is a ground-fixed coordinate system (also known as the Earth coordinate system) with the center of the earth as its origin.
  • ECEF coordinates The system is firmly connected to the earth and rotates with the earth.
  • the origin O(0,0,0) is the center of
  • X K is an ECEF coordinate system x-axis value
  • Y K is ECEF coordinate system y-axis values
  • Z K ECEF coordinate system is a z-axis value
  • N e is a front vertical radius of curvature, e is the eccentricity of the earth ellipsoid
  • a is the semi-major axis of the earth's ellipsoid, which is the radius of the earth's equator, which is 6378137 meters;
  • b is the semi-major axis of the earth's ellipsoid, which is the polar radius of the Earth, which is 635,6752.3 meters;
  • the origin is the center of mass of the earth, and the track is expressed as:
  • Heading It is used to describe the turning maneuvering characteristics of the sailing track on the apron surface.
  • the acceleration a k is used to describe the acceleration and deceleration movement characteristics of the sailing track on the apron scene, the formula is as follows:
  • V K and T K are respectively the "pass point speed” and "pass point time” in the voyage record information of the surface operation target.
  • the specific record information includes serial number, operation target type, model, task type number, label number, navigation Trace points, coordinate positions, passing point time, passing point speed, heading angle and acceleration, etc., are used as the basic data of the navigation motion intention recognition method.
  • Step 2) According to the apron scene site environment, construct the scene operation intention model of the site. Based on the establishment of the navigation track sample database of the moving target apron scene, associate the sample library with the scene operation intention model and mark the moving target The operation intention model category of the navigation track sample database of the apron scene is established, and the operation intention recognition model of the moving target of the airport apron scene is established; the specific steps are as follows:
  • the movement intent of the aircraft or vehicle is basically in the intersection area.
  • the position marked by the black dot in the figure is the "movement intention recognition zone";
  • the “movement intention recognition zone” is basically divided into three types: four forks, three forks and two forks.
  • Figure 2 shows the movement change intention of the moving target from the intersection R11 through the four-fork road R1.
  • I the road junction (or movement intention recognition area) number
  • I the operation intention type number
  • Table 3 is an example of the movement change intention model classification example of the moving target from the intersection R11 through the four-way road R1, as follows:
  • the scene operation intention model library which mainly contains three attributes, namely the operation intention model number, road junction, and operation intention model description.
  • the Reduce phase of the protocol model obtains the intermediate data of the Map, uses k-means (hard clustering algorithm), and calculates the heading angle and acceleration recognition range as the result information according to the type, model, task, movement intention, and position of the target.
  • k-means hard clustering algorithm
  • the recognition model mainly includes the operation target type, model, task, movement intention, position, and the recognition range of heading angle and acceleration, etc. .
  • the four situation identification diagrams are based on the acceleration a k model and the heading angle Model combination recognition.
  • the method of the present invention has been proved to be very effective through practice.
  • the method constructs a scene motion intention recognition model for aircraft and vehicles.
  • the model involves main key factors including target type, model, task attribute, motion intention, position, heading angle and acceleration recognition range, etc. , And use the motion intention recognition model of the present invention to correct the trajectory prediction in real time in real scenes, and achieve very good results, which lays the foundation for the research of new and comprehensive scene aircraft and vehicle trajectory prediction. It ensures the rapid and accurate intent recognition and trajectory operation prediction of the aircraft or vehicles that are active on the surface and about to move. While increasing the surface operating flow and reducing the workload of the controller, it effectively improves the level of airport traffic safety and efficiency. .

Abstract

一种基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,基于向量机统计分类理论实现运动目标雷达轨迹数据运行路线聚类,采用大数据Hadoop分布式运算架构,结合航空器或车辆的运动学模型,建立目标类型、型号、任务等属性特征和加速度、轨迹角、运动阶段运动特征相结合的目标特征集;通过对目标特征集和运行意图信息的离线训练与在线测试,开展运行目标场面滑行路线意图分析,构建机坪场面运动目标运行意图识别模型,提高了意图推理能力。

Description

基于雷达轨迹构建机坪场面运动目标运行意图识别的方法 技术领域
本发明属于民用航空空中交通管制(ATC)的机场机坪管制自动化技术领域,具体涉及一种基于雷达轨迹构建机坪场面运动目标运行意图识别的方法。
背景技术
我国航空运输业处于高速发展期,国内机场规模越来越大,逐步形成了多跑道运行、双塔台协同等局面,不断增长的航班量使得塔台管制的工作量逐年增加,多跑道运行也会增加地面引导难度。按照统计,中国民航的航空器起降次数从2003年的211.9万次增加到2015年的856.5万次,2015年的起降次数是2003年的4.04倍,机场场面发生不安全事件的风险随之变得越来越大。
就现在而言,国内一些大型机场装备高级场面活动引导与控制系统(A-SMGCS),具备监视、告警功能,一定程度上提高场面运行安全,但是由于缺乏对航空器或车辆在场面运行意图的预先识别,尤其大部分机场还不具备对车辆运行的监视和预测能力,场面冲突告警不能满足管制员提前预警的要求,因此避免飞机侵入运行跑道以及避免滑行道上各类运行冲突主要依靠管制员通过场面监视雷达及目视观察完成。
目前,对机场场面活动目标进行自动化运行监控已成为世界各国大型机场建设的主要目标之一。一方面,场面运行数据种类、接口方式众多,而且具有封闭性,安全性要求高;如何在不影响机场场面管制运行安全的情况下采集各类数据,体现数据价值成为重要的关键点。另一方面,运行数据体量非常庞大,数据分布特征多样;尤其雷达航迹数据,传统的系统架构和运算方法已经难以满足目标运行意图识别和活动预测相关应用的计算要求。
有鉴于此,本发明的方法结合航空器或车辆的动力和运动学模型,并通过基于Hadoop的海量雷达轨迹数据,分析航空器或车辆在场面活动全航行阶段中的速度、位置等信息,构建场面运动目标运行意图识别模型,并使用实际航迹数据进行实时修正,最后在实际工程项目中完成相关验证工作,为提高了场面活动目标运行预测能力打下基础,从而能够提前解决潜在的冲突,保障航班场面运行安全。
发明内容
针对于上述现有技术的不足,本发明的目的在于提供一种基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,以解决现有技术中依靠管制员对航空器或车辆运行意图识别、运动位置进行人工预测,很难区别实际环境下的各类航空器或车辆,及对运行意图识别和运动轨迹预测误差大,预测结果偏离高,数据可用性不强的问题。
为达到上述目的,本发明采用的技术方案如下:
本发明的一种基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,包括步骤如下:
步骤1):基于Hadoop的航空器或车辆运行目标特征集;
步骤2):根据机坪场面场地环境,构建该场地的场面运行意图模型,在运动目标机坪场面航行航迹样本库(简称航行航迹样本库)建立的基础上,把航行航迹样本库与场面运行意图模型进行关联,标注出航行航迹样本库所属运行意图模型类别,建立该机场机坪场面(简称本场)的运动目标运行意图识别模型。
进一步地,所述步骤1)具体包括:
11)对机坪场面航空器或车辆的航迹点记录的航迹数据做融合处理,处理后的数据与相应的运行路由计划信息匹配,建立场面运行目标航程记录信息;
12)利用任务类型号和标示号作为唯一性标识,按序列号排序相隔一个航迹点的两个航迹点航程记录信息构建数据文件的映射模型;
13)建立运动目标机坪场面航迹点之间的航向角算法模型;
14)保存Reduce(归约)阶段结果数据作为运动目标机坪场面航行航迹样本库,建立运动目标机坪场面运行航迹样本库。
进一步地,所述步骤11)中运行目标航程记录信息包括:运行目标类型(航空器或车辆)、型号(机型或车型)、任务类型号(航班号或任务单号)、标示号(机尾号/起落地时间或车牌号/任务发布时间)、航迹点、坐标位置、过点时间、过点速度。
需要说明的是,标示号:如航空器由“机尾号”和“起落地时间”拼建,如果为进港航班则采用落地时间,如果为出港航班则采用起飞时间;如车辆由“车牌号”和“任务发布时间”拼建。
以及,航迹数据和路由计划分为航空器和车辆两类;
一、航空器
航迹数据(多条数据):目标类型(航空器)、任务类型号(航班号)、航迹点、坐标位置、过点时间、过点速度;
路由计划(单条数据):目标类型(航空器)、型号(机型)、任务类型号(航班号)、机尾号、起落地时间、挡轮档时间(或撤轮档时间);
航迹数据和路由计划通过目标类型和任务类型号建立关联,如进港航班,则取航迹数据记录“过点时间”在路由计划记录“落地时间”和“挡轮档时间”之间数据;如出港,则取路由计划记录“撤轮档时间”和“起飞时间”之间数据。
二、车辆
航迹数据(多条数据):目标类型(车辆)、任务类型号(任务单号)、航迹点、坐标位置、过点时间、过点速度;
路由计划(单条数据):目标类型(车辆)、型号(车型)、任务类型号(任务单号)、车牌号、任务发布时间、任务结束时间;
航迹数据和路由计划通过目标类型和任务类型号建立关联,取航迹数据记录“过点时间” 在路由计划记录“任务发布时间”和“任务结束时间”之间数据。
进一步地,所述步骤13)具体包括:雷达航迹三维位置观测数据采用WGS-84坐标系,B K为WGS-84坐标系K航迹点的经度,L K为WGS-84坐标系K航迹点的纬度,H K为WGS-84坐标系K航迹点的高度;航迹表示为:
Traj K={B K,L K,H K},K=1,...,N
首先将WGS-84坐标系转换到地心地固直角坐标系ECEF,转换公式如下:
X K=(N e+H K)COS(L K)COS(B K)
Y K=(N e+H K)COS(L K)SIN(B K)
Z K=(N e(1-e 2)+H K)SIN(L K)
式中,X K是ECEF坐标系x轴值;Y K是ECEF坐标系y轴值;Z K是ECEF坐标系z轴值;N e是主垂直面的曲率半径,
Figure PCTCN2019110514-appb-000001
e是地球椭球偏心率,
Figure PCTCN2019110514-appb-000002
其中,a是地球椭球的长半轴,即地球赤道半径,a取6378137米;b是地球椭球的短半轴,即地球极半径,b取6356752.3米;
在ECEF坐标系下,原心为地球质心,航迹表示为:
Traj K={X K,Y K,Z K},K=1,...,N
然后,采用Traj K-1和Traj K+1航迹点的ECEF坐标位置、过点速度、过点时间计算Traj K航迹点的航向角
Figure PCTCN2019110514-appb-000003
和加速度a k,并作为Reduce阶段结果数据保存;
航向角
Figure PCTCN2019110514-appb-000004
用来描述航行航迹在机坪场面上转弯机动特征,公式如下:
Figure PCTCN2019110514-appb-000005
加速度a k用来描述航行航迹在机坪场面上加减速运动特征,公式如下:
Figure PCTCN2019110514-appb-000006
式中,V K和T K分别为场面运行目标航程记录信息中“过点速度”及“过点时间”。
进一步地,所述步骤2)具体包括:
21)在机场机坪场面道路图上,航空器或车辆的运动意图改变基本在交叉路口区域,以交叉点为中心,设定附近区域为运动意图识别区;
22)分析各类运动意图识别区的特点,对运动意图识别区的运行意图类型进行分类;
23)把航行航迹样本库、运动意图识别区与场面运行意图模型进行关联,标注出航行航迹样本库所属运行意图模型类别;
24)保存Reduce阶段结果数据作为经验数据模型,利用运行目标特征集、模拟或实时采录现场航迹运行数据去修正,同时结合航空器或车辆的运动学模型,通过运行意图信息的离线训练与在线测试开展自我学习,保证识别模型的完整性和唯一性,最终建立该机场机坪场面运动目标运行意图识别模型。
本发明的有益效果:
1、采用大数据分布式体系架构替代传统体系架构,解决了传统体系架构对海量数据难以运算的问题,高效的获得运算结果;
2、通过大数据支撑获得的识别模型替代人工经验,结合各种类型航空器和车辆运动学模型,细化了识别模型分类,减少了意图模型识别属性单一性,提高预测的精确性;
3、采用人工智能方法,根据实际数据对意图识别模型不断实时修正,进一步提升航行预测准确度,为下一工作提早做好规划,大大降低乃至避免场面冲突,提升场面运行安全。
4、高精度的意图识别和航行预测同时提高交通安全水平和效率水平,在一定程度上也会提升场面运行流量,减轻管制员的工作负荷,提升航空运输服务能力。
附图说明
图1为机场机坪场面道路图;
图2为动目标从路口R11经过四叉路R1的运行改变意图;
图3为示例中运行目标以加速度a k识别运行意图模型图;
图4为示例中运行目标以航向角
Figure PCTCN2019110514-appb-000007
识别运行意图模型图;
图5为地心地固直角坐标系图。
具体实施方式
本发明的方法结合航空器或车辆动力和运动学模型,并通过基于海量场面雷达轨迹数据,采用大数据Hadoop分布式运算框架,分析航空器或车辆在场面活动全航行阶段中的速度、位置等信息,建立目标类型、型号、任务等属性特征和加速度、轨迹角、运动阶段等运动特征相结合的目标特征集;通过对目标特征集和运行意图信息的离线训练与在线测试,开展运行目标场面滑行路线意图分析,构建机坪场面运动目标运行意图识别模型,模型涉及主要关键因子包括目标类型、型号、任务属性、运动意图、位置、航向角(即飞机和航天飞机的纵轴与地球北极之间的夹角)和加速度识别范围等,并在真实场景中使用本发明的运动意图识别 模型对轨迹预测进行实时修正,取得非常好的效果,为研究新型全面的场面航空器和车辆轨迹预测打下基础。这种推算的结果越准确,就越有可能将场面活动目标之间的冲突尽早探测出来,并加以调整消解,从而大大降低乃至避免场面冲突的可能性,提升场面运行安全;而另一方面,这种推算的结果越准确,就越有利于在当前时间点之前对所有场面运行情况进行总体把握,从而可以较早的平滑交通流,增加交通通过量,提升了交通效率。所以说,高精度的意图识别和航行预测是同时提高交通安全水平和效率水平的重要手段,在一定程度上也会提升场面运行流量,减轻管制员的工作负荷,产生较大的经济和社会效益。
为了便于本领域技术人员的理解,下面结合实施例与附图对本发明作进一步的说明,实施方式提及的内容并非对本发明的限定。
参照图1所示,本发明的一种基于机坪场面运动目标雷达轨迹构建目标运行意图识别模型的方法,包括步骤如下:
步骤1):基于Hadoop的航空器或车辆运行目标特征集;具体包括步骤如下:
11)对机坪场面航空器或车辆的航迹点记录的航迹数据作融合处理,清洗过后的数据与相应的运行路由计划信息匹配,建立机坪场面运行目标航程记录信息。在关系型数据库的海量场面运行目标滑行记录表中,每一条机坪场面航程记录由运行目标类型(航空器或车辆)、型号(机型或车型)、任务类型号(航班号或任务单号)、标示号(机尾号/起落地时间或车牌号/任务发布时间)、航迹点、坐标位置、过点时间、过点速度等属性组成。删除记录中有属性为空的无效记录;增加序列号字段,填写数值为此活动目标在场面运行过程中按过点时间排序。将序列号、运行目标类型、型号、任务类型号、标示号、航迹点、坐标位置、过点时间、过点速度九个维度的场面航程记录迁移到Hadoop集群的分布式数据库HBase中;
12)利用任务类型号和标示号作为唯一性标识,按序列号排序相隔一个航迹点的两个航迹点航程记录信息构建数据文件的映射模型;参见表1,其为HBase中场面航程记录信息,如下:
表1
Figure PCTCN2019110514-appb-000008
Figure PCTCN2019110514-appb-000009
在映射模型的Map阶段,Map过程将存储于HBase中的原始数据映射为间隔两个航迹点相关航程记录信息,Map的中间数据具体信息项如下表2;
表2
Figure PCTCN2019110514-appb-000010
Figure PCTCN2019110514-appb-000011
13)建立机坪场面航迹点之间的航向角算法模型;雷达航迹三维位置观测数据采用WGS-84坐标系(WGS-84坐标系是目前国际上统一采用的大地坐标系,GPS广播星历是以WGS-84坐标系为根据的),B K为WGS-84坐标系K航迹点的经度,L K为WGS-84坐标系K航迹点的纬度,H K为WGS-84坐标系K航迹点的高度。航迹表示为:
Traj K={B K,L K,H K},K=1,...,N
首先将WGS-84坐标系转换到地心地固直角坐标系ECEF,转换公式如下:
X K=(N e+H K)COS(L K)COS(B K)
Y K=(N e+H K)COS(L K)SIN(B K)
Z K=(N e(1-e 2)+H K)SIN(L K)
图5为地心地固直角坐标系(Earth-Centered,Earth-Fixed,简称ECEF)简称地心坐标系,是一种以地心为原点的地固坐标系(也称地球坐标系),ECEF坐标系与地球固联,且随着地球转动。原点O(0,0,0)为地球质心,z轴与地轴平行指向北极点,x轴指向本初子午线与赤道的交点,y轴垂直于xOz平面(即东经90度与赤道的交点)构成右手坐标系。
式中,X K是ECEF坐标系x轴值;Y K是ECEF坐标系y轴值;Z K是ECEF坐标系z轴值;N e是主垂直面的曲率半径,
Figure PCTCN2019110514-appb-000012
e是地球椭球偏心率,
Figure PCTCN2019110514-appb-000013
其中,a是地球椭球的长半轴,即地球赤道半径,取6378137米;b是地球椭球的短半轴,即地球极半径,取6356752.3米;
在ECEF坐标系下,原心为地球质心,航迹表示为:
Traj K={X K,Y K,Z K},K=1,...,N
然后,采用Traj K-1和Traj K+1航迹点的ECEF坐标位置、过点速度、过点时间计算Traj K航迹点的航向角
Figure PCTCN2019110514-appb-000014
和加速度a k,并作为Reduce阶段结果数据保存;
航向角
Figure PCTCN2019110514-appb-000015
用来描述航行航迹在机坪场面上转弯机动特征,公式如下:
Figure PCTCN2019110514-appb-000016
加速度a k用来描述航行航迹在机坪场面上加减速运动特征,公式如下:
Figure PCTCN2019110514-appb-000017
式中,V K和T K分别为场面运行目标航程记录信息中“过点速度”及“过点时间”。
14)保存Reduce阶段结果数据作为运动目标机坪场面航行航迹样本库,建立机坪场面运行航迹样本库,具体记录信息包括序列号、运行目标类型、型号、任务类型号、标示号、航迹点、坐标位置、过点时间、过点速度、航向角和加速度等,作为航行运动意图识别方法的基础数据。
步骤2):根据机坪场面场地环境,构建该场地的场面运行意图模型,在运动目标机坪场面航行航迹样本库建立的基础上,把样本库与场面运行意图模型进行关联,标注运动目标机坪场面航行航迹样本库所属运行意图模型类别,建立该机场机坪场面的运动目标运行意图识别模型;具体包括步骤如下:
21)在机场机坪场面道路图上,航空器或车辆的运动意图改变基本在交叉路口区域,我们以交叉点为中心,设定附近区域为“运动意图识别区”。参照图1,图中黑点标注的位置即为“运动意图识别区”;
梳理整个机场机坪场面道路图,建立运动意图识别区参数表,主要包含2个属性,分别为运动意图识别区号、区域范围。
22)分析各类“运动意图识别区”特点,开展“运动意图识别区”运行意图类型分类;
“运动意图识别区”即交叉路口的设置基本分为四叉、三叉和两叉三种类型。图2为动目标从路口R11经过四叉路R1的运行改变意图,运动目标运行到P0点时,会有四种运行改变情况,分别为停止、左转、直行和右转,记为运行意图P R,I,R为道路口(或运动意图识别区)编号,I为运行意图类型编号,表3为动目标从路口R11经过四叉路R1的运行改变意图模型分类示例,如下:
表3
Figure PCTCN2019110514-appb-000018
Figure PCTCN2019110514-appb-000019
因此,航空器或车辆分别从不同路口经过四叉路就会有16种运行意图模型。同理,路过三叉、两叉路口,分别有九种、四种运行改变情况。
根据机坪场面的“运动意图识别区”和运行意图类别,建立场面运行意图模型库,主要包含3个属性,分别为运行意图模型编号、道路口、运行意图模型描述。
23)把航行航迹样本库、运动意图识别区与场面运行意图模型进行关联,标注出航行航迹样本库所属运行意图模型类别,特别注意样本库中同一标示号航迹运行路线与运行意图模型类别向对应,即航迹路线从哪个路口到哪个路口;同时根据场面航行航迹样本数据模拟补充运行意图模型类别缺少的航迹样本数据,如在路口“运行停止”等情况。并把关联数据构建数据文件的映射模型,在映射模型的Map阶段,Map过程将存储于HBase中的原始数据映射为中间数据,中间数据包括目标类型、型号、任务类型号、航迹点、坐标位置、航向角、加速度、运动意图识别区号和运行意图模型编号。规约模型Reduce阶段获取Map的中间数据,使用k-means(硬聚类算法),按照运行目标类型、型号、任务、运动意图、位置计算出航向角和加速度识别范围作为结果信息。
24)保存Reduce阶段结果数据作为经验数据模型,利用运行目标特征集、模拟或实时采录现场航迹运行数据去修正,同时结合航空器或车辆的运动学模型,通过运行意图信息的离线训练与在线测试开展自我学习,保证识别模型的完整性和唯一性,最终建立本场运动目标运行意图识别模型,识别模型主要包括运行目标类型、型号、任务、运动意图、位置,以及航向角和加速度识别范围等。
示例:波音B777飞机进港航班SC1224由跑道滑行至停机位,经R11路口通过R1叉路的四种情况识别示意图,通过加速度a k模型和航向角
Figure PCTCN2019110514-appb-000020
模型结合识别。
“运行停止”P R11,1主要以通过加速度a k模型识别,示意图如下图3所示,当加速度a k在K线以下为“运行停止”P R11,1,K线以上为“左转”P R11,2、“直行”P R11,3和“右转”P R11,4
“左转”P R11,2、“直行”P R11,3和“右转”P R11,4主要以通过航向角
Figure PCTCN2019110514-appb-000021
模型识别,示意图如下图4,当航向角
Figure PCTCN2019110514-appb-000022
在线1和线2之间为“左转”P R11,2,线2和线3之间为“运行停止”P R11,1和“直行”P R11,3,线3和线4之间为“右转”P R11,4
本发明方法经实践使用证明非常有效,该方法构建了航空器和车辆的场面运动意图识别模型,模型涉及主要关键因子包括目标类型、型号、任务属性、运动意图、位置、航向角和加速度识别范围等,并在真实场景中使用本发明的运动意图识别模型对轨迹预测进行实时修正,取得非常好的效果,为研究新型全面的场面航空器和车辆轨迹预测打下基础。确保了对 正在场面活动的和即将活动的航空器或车辆快速准确意图识别、航迹运行预测,在提升场面运行流量、减轻管制员的工作负荷的同时,有效的提升机场场面交通安全水平和效率水平。
本发明具体应用途径很多,以上所述仅是本发明的优选实施方式,应当指出,对于本技术领域的普通技术人员来说,在不脱离本发明原理的前提下,还可以作出若干改进,这些改进也应视为本发明的保护范围。

Claims (5)

  1. 一种基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,其特征在于,包括步骤如下:
    步骤1):基于Hadoop的航空器或车辆运行目标特征集;
    步骤2):根据机坪场面场地环境,构建该场地的场面运行意图模型,在运动目标机坪场面航行航迹样本库建立的基础上,把航行航迹样本库与场面运行意图模型进行关联,标注出航行航迹样本库所属运行意图模型类别,建立该机场机坪场面的运动目标运行意图识别模型。
  2. 根据权利要求1所述的基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,其特征在于,所述步骤1)具体包括:
    11)对机坪场面航空器或车辆的航迹点记录的航迹数据做融合处理,处理后的数据与相应的运行路由计划信息匹配,建立场面运行目标航程记录信息;
    12)利用任务类型号和标示号作为唯一性标识,按序列号排序相隔一个航迹点的两个航迹点航程记录信息构建数据文件的映射模型;
    13)建立运动目标机坪场面航迹点之间的航向角算法模型;
    14)保存Reduce阶段结果数据作为运动目标机坪场面航行航迹样本库,建立运动目标机坪场面运行航迹样本库。
  3. 根据权利要求2所述的基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,其特征在于,所述步骤11)中运行目标航程记录信息包括:运行目标类型、型号、任务类型号、标示号、航迹点、坐标位置、过点时间、过点速度。
  4. 根据权利要求2所述的基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,其特征在于,所述步骤13)具体包括:雷达航迹三维位置观测数据采用WGS-84坐标系,B K为WGS-84坐标系K航迹点的经度,L K为WGS-84坐标系K航迹点的纬度,H K为WGS-84坐标系K航迹点的高度;航迹表示为:
    Traj K={B K,L K,H K},K=1,...,N
    首先将WGS-84坐标系转换到地心地固直角坐标系ECEF,转换公式如下:
    X K=(N e+H K)COS(L k)COS(B K)
    Y K=(N e+H K)COS(L K)SIN(B K)
    Z K=(N e(1-e 2)+H K)SIN(L K)
    式中,X K是ECEF坐标系x轴值;Y K是ECEF坐标系y轴值;Z K是ECEF坐标系z轴值; N e是主垂直面的曲率半径,
    Figure PCTCN2019110514-appb-100001
    e是地球椭球偏心率,
    Figure PCTCN2019110514-appb-100002
    a是地球椭球的长半轴,即地球赤道半径,b是地球椭球的短半轴,即地球极半径;
    在ECEF坐标系下,原心为地球质心,航迹表示为:
    Traj K={X K,Y K,Z K},K=1,...,N
    然后,采用Traj K-1和Traj K+1航迹点的ECEF坐标位置、过点速度、过点时间计算Traj K航迹点的航向角
    Figure PCTCN2019110514-appb-100003
    和加速度a k,并作为Reduce阶段结果数据保存;
    航向角
    Figure PCTCN2019110514-appb-100004
    用来描述航行航迹在机坪场面上转弯机动特征,公式如下:
    Figure PCTCN2019110514-appb-100005
    加速度a k用来描述航行航迹在机坪场面上加减速运动特征,公式如下:
    Figure PCTCN2019110514-appb-100006
    式中,V K和T K分别为场面运行目标航程记录信息中过点速度及过点时间。
  5. 根据权利要求1所述的基于雷达轨迹构建机坪场面运动目标运行意图识别的方法,其特征在于,所述步骤2)具体包括:
    21)以交叉点为中心,设定附近区域为运动意图识别区;
    22)分析各类运动意图识别区的特点,对运动意图识别区的运行意图类型进行分类;
    23)把航行航迹样本库、运动意图识别区与场面运行意图模型进行关联,标注出航行航迹样本库所属运行意图模型类别;
    24)保存Reduce阶段结果数据作为经验数据模型,利用运行目标特征集、模拟或实时采录现场航迹运行数据去修正,同时结合航空器或车辆的运动学模型,通过运行意图信息的离线训练与在线测试开展自我学习,保证识别模型的完整性和唯一性,最终建立该机场机坪场面运动目标运行意图识别模型。
PCT/CN2019/110514 2019-05-15 2019-10-11 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法 WO2020228228A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910401652.9A CN110111608B (zh) 2019-05-15 2019-05-15 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法
CN201910401652.9 2019-05-15

Publications (1)

Publication Number Publication Date
WO2020228228A1 true WO2020228228A1 (zh) 2020-11-19

Family

ID=67490145

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/110514 WO2020228228A1 (zh) 2019-05-15 2019-10-11 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法

Country Status (2)

Country Link
CN (1) CN110111608B (zh)
WO (1) WO2020228228A1 (zh)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113625236A (zh) * 2021-06-30 2021-11-09 嘉兴聚速电子技术有限公司 多雷达数据融合方法、装置、存储介质和设备
CN113987036A (zh) * 2021-12-03 2022-01-28 南京莱斯信息技术股份有限公司 一种空管实时数据转换成模拟训练场景的方法
CN114049602A (zh) * 2021-10-29 2022-02-15 哈尔滨工业大学 一种基于意图推理的逃逸目标跟踪方法及系统
CN114187783A (zh) * 2021-12-06 2022-03-15 中国民航大学 一种用于机场飞行区潜在冲突分析预测的方法
CN114783216A (zh) * 2022-04-11 2022-07-22 南京莱斯信息技术股份有限公司 一种机场场面数字化引导系统及方法
CN115983613A (zh) * 2023-03-21 2023-04-18 民航成都信息技术有限公司 一种航班拼接方法、装置、电子设备及存储介质
CN117076839A (zh) * 2023-10-17 2023-11-17 中国民用航空总局第二研究所 基于双重增量神经网络的机场航空器轨迹动态预测方法

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110111608B (zh) * 2019-05-15 2021-06-18 南京莱斯信息技术股份有限公司 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法
CN110364034A (zh) * 2019-08-26 2019-10-22 中国航空工业集团公司沈阳飞机设计研究所 一种无人机滑行通道描述方法及系统
CN114783214A (zh) * 2020-04-28 2022-07-22 上海波若智能科技有限公司 路网动态数据采集方法及路网动态数据采集系统
CN113443296A (zh) * 2020-08-17 2021-09-28 刘炎 基于大数据的垃圾自动分类系统
CN112308171A (zh) * 2020-11-23 2021-02-02 浙江天行健智能科技有限公司 一种基于模拟驾驶器的车辆位置预测建模方法
CN114358211B (zh) * 2022-01-14 2022-08-23 中科世通亨奇(北京)科技有限公司 基于多模态深度学习的航空器行为意图识别方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103093649A (zh) * 2011-10-31 2013-05-08 通用电气公司 用于推断飞行器参数的方法和系统
CN104717468A (zh) * 2015-03-06 2015-06-17 中国科学院大学 基于集群轨迹分类的集群场景智能监控方法及系统
CN108268597A (zh) * 2017-12-18 2018-07-10 中国电子科技集团公司第二十八研究所 一种动目标活动概率图谱构建及行为意图识别方法
US20190005744A1 (en) * 2013-04-22 2019-01-03 Latitude Technologies Corporation Aircraft flight data monitoring and reporting system and use thereof
CN109587655A (zh) * 2017-09-29 2019-04-05 波音公司 用于通过标准飞机轨迹公约来传送高保真飞机轨迹相关信息的系统和方法
CN110111608A (zh) * 2019-05-15 2019-08-09 南京莱斯信息技术股份有限公司 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101655561A (zh) * 2009-09-14 2010-02-24 南京莱斯信息技术股份有限公司 基于联合卡尔曼滤波的多点定位数据与雷达数据融合方法
CN103592949B (zh) * 2013-11-28 2016-03-02 电子科技大学 一种用于无人机编队同时到达目标的分布式控制方法
FR3034859B1 (fr) * 2015-04-10 2018-08-17 Airbus Operations (S.A.S.) Dispositif, systeme et procede d'aide au roulage au sol d'un aeronef
CN105812733B (zh) * 2016-03-15 2017-12-15 中国民用航空总局第二研究所 一种民航空中交通管制的场面监视引导系统
US9969386B1 (en) * 2017-01-10 2018-05-15 Mitsubishi Electric Research Laboratories, Inc. Vehicle automated parking system and method
CN208256104U (zh) * 2018-01-23 2018-12-18 上海空地信息技术有限公司 一种基于ads-b的通航机场场面监视系统
CN109188423B (zh) * 2018-08-29 2020-11-10 电子科技大学 一种基于多源聚类的分布式多目标跟踪方法
CN109598983A (zh) * 2018-12-12 2019-04-09 中国民用航空飞行学院 一种机场场面光电监视告警系统及方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103093649A (zh) * 2011-10-31 2013-05-08 通用电气公司 用于推断飞行器参数的方法和系统
US20190005744A1 (en) * 2013-04-22 2019-01-03 Latitude Technologies Corporation Aircraft flight data monitoring and reporting system and use thereof
CN104717468A (zh) * 2015-03-06 2015-06-17 中国科学院大学 基于集群轨迹分类的集群场景智能监控方法及系统
CN109587655A (zh) * 2017-09-29 2019-04-05 波音公司 用于通过标准飞机轨迹公约来传送高保真飞机轨迹相关信息的系统和方法
CN108268597A (zh) * 2017-12-18 2018-07-10 中国电子科技集团公司第二十八研究所 一种动目标活动概率图谱构建及行为意图识别方法
CN110111608A (zh) * 2019-05-15 2019-08-09 南京莱斯信息技术股份有限公司 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113625236A (zh) * 2021-06-30 2021-11-09 嘉兴聚速电子技术有限公司 多雷达数据融合方法、装置、存储介质和设备
CN114049602A (zh) * 2021-10-29 2022-02-15 哈尔滨工业大学 一种基于意图推理的逃逸目标跟踪方法及系统
CN114049602B (zh) * 2021-10-29 2022-06-14 哈尔滨工业大学 一种基于意图推理的逃逸目标跟踪方法及系统
CN113987036A (zh) * 2021-12-03 2022-01-28 南京莱斯信息技术股份有限公司 一种空管实时数据转换成模拟训练场景的方法
CN113987036B (zh) * 2021-12-03 2024-04-09 南京莱斯信息技术股份有限公司 一种空管实时数据转换成模拟训练场景的方法
CN114187783B (zh) * 2021-12-06 2023-10-31 中国民航大学 一种用于机场飞行区潜在冲突分析预测的方法
CN114187783A (zh) * 2021-12-06 2022-03-15 中国民航大学 一种用于机场飞行区潜在冲突分析预测的方法
CN114783216B (zh) * 2022-04-11 2023-08-15 南京莱斯信息技术股份有限公司 一种机场场面数字化引导系统及方法
CN114783216A (zh) * 2022-04-11 2022-07-22 南京莱斯信息技术股份有限公司 一种机场场面数字化引导系统及方法
CN115983613A (zh) * 2023-03-21 2023-04-18 民航成都信息技术有限公司 一种航班拼接方法、装置、电子设备及存储介质
CN115983613B (zh) * 2023-03-21 2023-06-16 民航成都信息技术有限公司 一种航班拼接方法、装置、电子设备及存储介质
CN117076839A (zh) * 2023-10-17 2023-11-17 中国民用航空总局第二研究所 基于双重增量神经网络的机场航空器轨迹动态预测方法
CN117076839B (zh) * 2023-10-17 2023-12-26 中国民用航空总局第二研究所 基于双重增量神经网络的机场航空器轨迹动态预测方法

Also Published As

Publication number Publication date
CN110111608B (zh) 2021-06-18
CN110111608A (zh) 2019-08-09

Similar Documents

Publication Publication Date Title
WO2020228228A1 (zh) 基于雷达轨迹构建机坪场面运动目标运行意图识别的方法
CN108196575B (zh) 一种无人机任务分配及路线规划方法
Kanistras et al. A survey of unmanned aerial vehicles (UAVs) for traffic monitoring
CN104819724B (zh) 一种基于gis的无人地面车辆自主行驶辅助系统
Patron et al. Flight trajectories optimization under the influence of winds using genetic algorithms
CN105303896B (zh) 航班预计到达时间精确预估方法
CN112116830A (zh) 一种基于空域网格化的无人机动态地理围栏规划方法
CN109493644A (zh) 一种基于历史航迹数据挖掘的四维航迹推测方法
CN103984355B (zh) 一种巡检飞行机器人与架空电力线路距离预测和保持方法
CN111009155B (zh) 基于空域结构和航班流的空中交通流复杂性量化分析方法
Zhang et al. Efficient vehicles path planning algorithm based on taxi GPS big data
CN112735189A (zh) 一种飞行汽车地空模态转换与智能空管方法及系统
CN109101743A (zh) 一种高精度路网模型的构建方法
CN103678925B (zh) 基于辅助信源的航迹分类方法
Wang et al. Airport taxiway conflict detection method based on network topology
US11688291B2 (en) Cockpit display systems and methods for displaying taxiing route on airport moving map
Jiang et al. Research on method of trajectory prediction in aircraft flight based on aircraft performance and historical track data
Ni et al. Design of regional logistics system based on unmanned aerial vehicle
Shangguan et al. Interactive perception-based multiple object tracking via CVIS and AV
CN105096661B (zh) 基于垂直剖面视图的空中交通流管理系统与方法
Li et al. Autonomous navigation technology for low-speed small unmanned vehicle: An overview
Ziarmand et al. Transport monitoring and control systems
Tyagi et al. Intelligent transportation system: Need, working, and tools
Wang et al. Lightweight High-precision Map for Specific Scenes
Wang et al. Design of UAV's safe obstacle avoidance trajectory based on point cloud

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19929094

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19929094

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205N DATED 14/01/2022)

122 Ep: pct application non-entry in european phase

Ref document number: 19929094

Country of ref document: EP

Kind code of ref document: A1