Title :
Study of rapid collision detection algorithm for manipulator
Author :
Yue Shen;Qingxuan Jia;Gang Chen;Yifan Wang;Hanxu Sun
Author_Institution :
School of Automation, Beijing University of Posts and Telecommunications, China
fDate :
6/1/2015 12:00:00 AM
Abstract :
Bounding box algorithm is generally used to detect whether a manipulator is in collision with environment or itself. In this paper, we propose a rapid collision detection algorithm for manipulator based on the vector relation of basic geometrical elements. Firstly, we analyze the precision and efficiency of three commonly used algorithm - sphere, AABB (Axis-Aligned Bounding Box) and OBB(Oriented Bounding Box). Secondly, we extract the concept of “center distance” and “geometric dimension” by simplifying a manipulator and its environment to the combination of some geometry for detecting collision rapidly. Then, we simplify the collision of manipulator to 4 kinds of collision - sphere-cylinder, sphere-cuboid, cylinder-cylinder and cuboid-cuboid. Finally, the validity of this collision detection algorithm is verified by developing a simulation system based on OSG(OpenSceneGraph) for a 8-DOF manipulator.
Keywords :
"Manipulators","Collision avoidance","Geometry","Algorithm design and analysis","Image edge detection","Mathematical model","Shape"
Conference_Titel :
Industrial Electronics and Applications (ICIEA), 2015 IEEE 10th Conference on
DOI :
10.1109/ICIEA.2015.7334244