Autonomous Navigation with Obstacle Avoidance of a Nonholonomic Four Wheel Steering Robotic Platform
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
INTERACT is a space technology demonstration experiment in which an advanced wheeled mobile robot (WMR) will be tele-operated from space. The experiment consists of a navigation phase and a manipulation phase. To cope with adverse communication conditions and limited means for control, an autonomous navigation algorithm needs to be developed for the navigation phase of the experiment. The algorithm requires accurate localization, path planning among obstacles and feedback control of the nonholonomic system. The main goal is to autonomously position the platform in front of a taskboard within 15 cm in position and 6º in heading. Such a navigational algorithm has not been developed before for a four wheel steering platform capable of crab motion and rotation in place. Furthermore, it is a challenge to achieve the necessary localization accuracy with the available sensors. The navigation algorithm is implemented on the actual INTERACT platform and its performance is experimentally validated. The algorithm is limited to 2D planning and control. The path is planned using the Rapid-exploration Random Tree algorithm and tracking control is performed using approximate linearization around the reference trajectory and PID control laws. Feedback is provided by localization based on dead-reckoning and odometry. The experimental results show an average error over 10 trials of 14 cm in x, 13 cm in y and 4.7 degrees in heading, which is within limits.