题目：3D LiDAR Aided (3DLA) Urban GNSS Positioning for Autonomous Driving
时间：2023年3月4日 10:30-11:30
地点：机械与动力工程学院 F310会议室
报告人：Dr. Weisong Wen（香港理工大学）
邀请人：王亚飞 副教授（智能汽车研究所）
Biography
Dr. Weisong Wen is a Research Assistant Professor at the De
partment of Aeronautical and Aviation Engineering, The Hong Kong Polytechnic University. He was the interim Director of Intelligent Positioning and Navigation Lab-IPNL. Dr. Wen received a Ph.D. degree in Mechanical Engineering from The Hong Kong Polytechnic University (PolyU)， and he was also a visiting Ph.D. student with the Faculty of Engineering, University of California, Berkeley (UC Berkeley) in 2018. Before joining PolyU as a Research Assistant Professor, he was a senior research fellow at AAE of PolyU. Dr. Wen has secured more than 4.5 million in research funds from both academics and industry as the role of PI.
Abstract
Positioning in urban environments is becoming essential due to the increasing demand for autonomous driving vehicles (ADV). The global navigation satellite system (GNSS) is currently one of the principals means of providing globally-referenced positioning for ADV localization. The GNSS positioning accuracy is significantly degraded in highly-urbanized cities such as Hong Kong, due to signal reflection caused by static buildings and dynamic objects such as double-decker buses. If the direct light-of-sight (LOS) is blocked, and reflected signals from the same satellite are received, the notorious non-light-of-sight (NLOS) receptions occur. Inspired by the strong perception capability of ADV using onboard sensors (such as 3D LiDAR), we continuously developed perception-aided NLOS mitigation methods where the 3D LiDAR is employed to timely reconstruct the surrounding environments to identify the NLOS receptions. The idea was also reported in the industrial magazine in 2018. The work was further improved in 2020, where several drawbacks are relaxed and were awarded the Best Presentation Award in the session on Navigation in Urban Environments.
