Velocity obstacle based local collision avoidance for a holonomic elliptic robot

• Published in: Autonomous robots Vol. 41; no. 6; pp. 1347 - 1363
• Main Authors: Lee, Beom H., Jeon, Jae D., Oh, Jung H.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.08.2017; Springer Nature B.V
• Subjects: Algorithms; Angular velocity; Artificial Intelligence; Collision avoidance; Collisions; Computer Imaging; Computer simulation; Control; Engineering; Mechatronics; Motion planning; Pattern Recognition and Graphics; Robotics; Robotics and Automation; Robots; Rotation; Travel time; Velocity; Vision; Motion planning; Holonomic robot; Ellipse; Collision avoidance; Dynamic environment
• ISSN: 0929-5593; 1573-7527
• DOI: 10.1007/s10514-016-9580-2

This paper addresses the local collision avoidance problem for a holonomic elliptic robot, where its footprint and obstacles are approximated with the minimum area bounding ellipses. The proposed algorithm is decomposed into two phases: linear and angular motion planning. In the former phase, the ellipse-based velocity obstacle is defined as a set of all linear velocities of the robot that would cause a collision with an obstacle within a finite time horizon. If the robot’s new linear velocity is selected outside of the velocity obstacle, the robot can avoid the obstacle without rotation. In the latter phase, the angular velocity is selected at which the robot can circumvent the obstacle with the minimum possible deviation by finding the collision-free rotation angles and the preferred angular velocities. Finally, the performance of the suggested algorithm is demonstrated in simulation for various scenarios in terms of travel time, distance, and the number of collisions.