DocumentCode :
1868090
Title :
Trajectory inverse kinematics by conditional density modes
Author :
Qin, Chao ; Carreira-Perpinan, Miguel A.
Author_Institution :
Sch. of Eng., Univ. of California at Merced, Merced, CA
fYear :
2008
fDate :
19-23 May 2008
Firstpage :
1979
Lastpage :
1986
Abstract :
We present a machine learning approach for trajectory inverse kinematics: given a trajectory in workspace, to find a feasible trajectory in angle space. The method learns offline a conditional density model of the joint angles given the workspace coordinates. This density implicitly defines the multivalued inverse kinematics mapping for any workspace point. At run time, given a trajectory in the workspace, the method (1) computes the modes of the conditional density given each of the workspace points, and (2) finds the reconstructed angle trajectory by minimising over the set of modes a global, trajectory-wide constraint that penalises discontinuous jumps in angle space or invalid inverses. We demonstrate the method with a PUMA 560 robot arm and show how it can reconstruct the true angle trajectory even when the workspace trajectory contains singularities, and when the number of inverse branches depends on the workspace location.
Keywords :
kinematics; learning (artificial intelligence); PUMA 560 robot arm; angle trajectory; conditional density model; conditional density modes; machine learning; multivalued inverse kinematics mapping; trajectory inverse kinematics; workspace coordinates; Boundary conditions; Computational efficiency; Equations; Jacobian matrices; Machine learning; Neural networks; Optimization methods; Robot kinematics; Robotics and automation; Sequences;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Robotics and Automation, 2008. ICRA 2008. IEEE International Conference on
Conference_Location :
Pasadena, CA
ISSN :
1050-4729
Print_ISBN :
978-1-4244-1646-2
Electronic_ISBN :
1050-4729
Type :
conf
DOI :
10.1109/ROBOT.2008.4543497
Filename :
4543497
Link To Document :
بازگشت