Mobile Robot Localization and Trajectory Optimization

Localization;Nonlinear Optimization;Nonholonomic;Mobile Robot;Mechanical Engineering

Xu, Ying

Johns Hopkins University

关键词: Localization;    Nonlinear Optimization;    Nonholonomic;    Mobile Robot;    Mechanical Engineering;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/37301/rustler.avi?sequence=2&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

Autonomous Mobile robot navigation has become a popular topic in robotics due to its emerging applications in self-driving vehicles and autonomous air-drones. This essay explores the two main components of navigation namely perception and control. Simultaneous localization and mapping (SLAM) has been one of the most widely adoptedcollection of methods for self localization of robots in unknown environments. Here we put our focus on gMapping for offline map building and Adaptive Monte Carlo Localization (AMCL) for realtime localization. On the control side, Forward Backward Sweep Method is used to generate locally optimal trajectories and the corresponding feedback control law. Our experiments show that using the above methods and a properly integrated system autonomous navigation can be achieved with up to 2 m/s navigation speed.