1981 Volume 17 Issue 5 Pages 582-588
In this paper we consider the dynamic control of manipulators from the view point of mechanics, and propose a new method for task-oriented coordinate control which does not require a transformation from task-oriented coordinates to joint angles. The proposed method can be easily implemented by a microcomputer and suited to the sensor feedback control. Furthermore, it is shown that the proposed method is applicable even when holonomic constraints are added to the system, and when redundancy or singularity arises in the relation between task-oriented coordinates and joint angles. Effectiveness of the proposed method is verified by computer simulation utilizing an arm model with four degrees of freedom.