ABSTRACT

The key step in computed torque control is that the motion is completely specified by the desired position, velocity and acceleration, of all the joints of a robot manipulator. Thus, the first issue that must be considered is the efficient specification of the desired position, velocity and acceleration, of all the joints of a robot manipulator. This is the path planning problem and the paths may be specified either in the Cartesian coordinates or in the joint space. The problem with path planning in the joint space is the need to satisfy constraints, such as maintaining constant velocity, in the Cartesian space. The problem is related to the issue of dynamic motion planning, which would become considerably easier if one could take into account the dynamics of the robot manipulator.