编者荐语：
原题：基于NDT扫描匹配之光达同步定位与地图构建演算法于定位、导航与授時之性能分析
以下文章来源于卫星导航国际期刊 ，作者邱玉婷
标题：基于NDT扫描匹配之光达同步定位与地图构建演算法于定位、导航与授时之性能分析
作者：江凯伟*，邱玉婷，Surachet Srinara，蔡孟伦
主题词：三维常态分布转换(3D NDT)；误差侦测与排除机制(FDE)；惯性与卫星导航整合系统(INS/GNSS)；LiDAR；定位、导航与授时(PNT)；同步定位与地图构建(SLAM)
Performance of LiDAR-SLAM-based PNT with initial poses based on NDT scan matching algorithm
Kaiwei Chiang*, Yuting Chiu, Surachet Srinara and Menglun Tsai
Satellite Navigation (2023) 4: 3
引用文章:
Chiang, K. W., Chiu, Y. T., Srinara, S. et al. Performance of LiDAR-SLAM-based PNT with initial poses based on NDT scan matching algorithm. Satell Navig 4, 3 (2023).
https://doi.org/10.1186/s43020-022-00092-0
PDF文件下载链接：
https://satellite-navigation.springeropen.com/articles/10.1186/s43020-022-00092-0
Editorial Summary
LiDAR-SLAM-based algorithm with FDE mechanisms for PNT application
To reach where-in-lane level accuracy (less than 0.5 m) for autonomous vehicular in positioning, navigation, and timing application. The authors raised a multi-sensor integration scheme by integrating LiDAR with GNSS/ INS through extended Kalman filter. However, considering without the assistance from high definition point cloud map, the quality of the received-data may be affected as the system operates in the diversity scenarios. Hence, this paper conducted fault detection and exclusion schemes for motion constraint and robust data extraction. After direct georeferencing, it can update the position, velocity and heading information by Normal Distribution Transform registration. To verify the algorithm, the result of the test system is compared to the reference system from navigation grade IMU. To sum up, the navigation scheme conducted in this paper can effectively reach navigation accuracy level and construct the dynamic point cloud map simultaneously.
本文亮点
1.惯性、卫星与光达整合导航系统于定位、导航与授時之应用。
2.基于NDT扫描匹配之光达同步定位与地图构建演算法。
3.误差侦测与排除机制之运用。
内容简介
为了提供自驾车达到车道中等级的定位精度（小于50公分），以作为定位、导航与授时之应用。本文在运用全球导航卫星系统与惯性导航系统的基础之上，藉由扩展式卡尔曼滤波器与光达进行多感测器整合。然而考量到在沒有高精点云地图的辅助下，系统于多样性的环境中运行将影响感测器接收的资料品质。作者提出误差侦测与排除机制以进行运动形态约制，并从中萃取可信的资料透过直接地理定位以及常态分布转换的方式进行扫描匹配以更新位置、速度、与姿态等导航资讯，误差侦测与排除机制将同步进行误差排除。为了验证本案提出的多感测器整合系统之导航能力，测试系统与导航等级的参考系统于多样态的环境进行测试与精度分析。综观而言，本文提出的导航系统架构将有效达到导航精度要求并能同步建构动态点云地图。
图文导读
图 1 应用于定位、导航与授时之基于光达同步定位与地图构建演算法架构图。
图 2 产制基于卫星导航系统定位解之方法示意图。(a) 差分全球导航卫星系统之概念示意图；(b) 航向角之估算方法示意图。
图 3 各感测器坐标系统间之转换关系示意图。
图 4 常态分布转换之概念示意图。
图 5 误差侦测与排除机制之应用。
图 6 感测器安装于载具示意图。(a) 感测器安装之俯视图；(b) 感测器安裝之室內图。
图 7 开阔区之测试场域。
图 8 卫星讯号干扰区之测试场域。
图 9 实验过程记录之位置精度衰減因子。
图 10 于开阔测试场域建构之动态更新地图。(a) 基于惯性与卫星导航整合系統之动态更新地图-俯视图；(b) 基于卫星导航系統之动态更新地图-俯视图；(c) 基于惯性与卫星导航整合系统之动态更新地图-侧视图；(d) 基于卫星导航系统之动态更新地图-侧视图。
图 11 本文提出之多感测器整合系统于开阔测试场域估算之轨迹成果。(a) 基于惯性与卫星导航整合定位解作为系统初始估算值；(b) 基于卫星导航系統之定位解作为系统初始估算值。
图 12 于卫星讯号干扰测试场域建构之动态更新地图。(a) 基于惯性与卫星导航整合系统之动态更新地图-俯视图；(b) 基于卫星导航系统之动态更新地图-俯视图；(c) 基于惯性与卫星导航整合系统之动态更新地图-侧视图；(d) 基于卫星导航系统之动态更新地图-侧视图。
图 13 本文提出之多感测器整合系统于卫星讯号干扰测试场域估算之轨迹成果。(a) 基于惯性与卫星导航整合定位解作为系统初始估算值；(b) 基于卫星导航系统之定位解作为系统初始估算值。
作者简介
江凯伟 教授
本文第一和通讯作者
台湾成功大学，中国台湾
▍作者简介
Dr. Kaiwei Chiang graduated from Department of Geomatics Engineering, University of Calgary with a doctorate degree. He is the professor of Department of Geomatics, Taiwan Cheng Kung University, also serves as the director of High Definition Maps Research Center. His research expertise includes Inertial Navigation Systems, Multi-sensor Fusion, and Optimal Estimation for Seamless Navigation and Mapping, and following are his research activities and achievements:
1.Seamless Navigation Engine Design for Lane Level Autonomous Vehicular Applications Using Automotive Grade Sensor Clusters.
2.Automated HD Maps Production Technologies Using Professional MMS and Certified 3rd-party Sourcing Measurements.
3.General Visual Odometry Aided Inertial Navigation Algorithms for Vehicular and Pedestrian Navigation and Mapping Applications.
4.Smart and Collaborative Mobile Mapping Platforms.
5.Optimal Sensor Fusion Strategies: Conventional and Artificial Intelligence.
Dr. Kaiwei Chiang published over 130 papers in academic journals, conference and workshop proceedings, and has been cited by 1183 literatures.
转自：“测绘学术资讯”微信公众号
如有侵权，请联系本站删除！
