An Outlier-Robust GNSS-inertial-LiDAR Localization System

Siwei Zhong, Chao Wei*, Jibin Hu, Ting Zhang, Jie Yu, Yongdan Chen

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

1 引用 (Scopus)

摘要

We tackle a modified Outlier-Robust GNSS-inertial-LiDAR unmanned vehicle localization system based on the factor graph. The frame of the localization system utilizes graph optimization to fuse information from IMU pre-integration, GNSS and LiDAR-inertial odometry. In order to cope with the problem of high-precision localization in complex scenes in driving process of the unmanned vehicle, residual x2 outlier test added before graph optimization applies in this frame to effectively eliminate outliers from GNSS and LiDAR-inertial odometry and mitigate their influence to maintain robust localization. In addition, a fixed-time sliding window is organized in optimization to lower the computation, satisfying real-time requirements. Through extensive experiments in simulations, the results show that this system can provide a reliable localization result and takes advantage over Kalman filter and pure LiDAR algorithm.

源语言英语
主期刊名ICSMD 2021 - 2nd International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence
出版商Institute of Electrical and Electronics Engineers Inc.
ISBN(电子版)9781665427470
DOI
出版状态已出版 - 2021
活动2nd International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence, ICSMD 2021 - Nanjing, 中国
期限: 21 10月 202123 10月 2021

出版系列

姓名ICSMD 2021 - 2nd International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence

会议

会议2nd International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence, ICSMD 2021
国家/地区中国
Nanjing
时期21/10/2123/10/21

指纹

探究 'An Outlier-Robust GNSS-inertial-LiDAR Localization System' 的科研主题。它们共同构成独一无二的指纹。

引用此