Applying 3D Lidar Odometry and Mapping Technique for Campus Autonomous Vehicle Navigation System Development

Autor: Ming-Liang Gao, 高銘良
Rok vydání: 2019
Druh dokumentu: 學位論文 ; thesis
Popis: 107
In this thesis, a campus autonomous vehicle project is proposed. The proposed system is divided into two parts: autonomous electric vehicle design and autopilot system. The front wheel steering and suspension system was done in terms of kinematics evaluation based on different dimension combination to meet minimum error of Ackermann steering geometry. This work used a front-wheel-steering with rear-wheel-drive configuration. To perform autonomous driving, a 3D LiDAR and a 2D LiDAR sensors were used for the map creation, localization and obstacle avoidance. In addition, the autopilot system also used the A* algorithm to realize route planning in our campus. The map creation and localization were carried out with the We use LiDAR odometry and mapping (LOAM). Meanwhile, the 2D LiDAR was used to detect obstacle in front of the vehicle so that the autonomous vehicle could properly avoid obstacles and pedestrians in from of the vehicle by either stopping or changing its driving trajectory according to the obstacle location. Finally, this study accomplished a bus route operation prototype system in our campus. The autonomous electric vehicle was able to depart and stop at the desired stops. Finally, the proposed autonomous electric vehicle system is a level 3 system defined by the society of automotive engineers (SAE); that means the drivers may take their hands off the wheel and feet off the pedals based on some specific situations with a standby driver ready for intervention when needed.
Databáze: Networked Digital Library of Theses & Dissertations