Hvbrid active/passive interaction control for robot manipulators

Abstract

Motion control of robot manipulators is typically accomplished by sensing and regulating the positions of the joint actuators to cause the end-effector to follow a programmed trajectory. Unfortunately, position control schemes usually fail for tasks that require significant physical interaction between the robot and a workpiece. This restriction severely limits the scope of work that robots are able to do. A solution to this problem is to incorporate some form of compliance into the manipulator to compensate for positioning errors. Researchers have been focused on two principal methods of achieving manipulator compliance - passive mechanism synthesis and active force control - both of which have advantages and disadvantages. In this paper, the authors present recent advances in passive mechanism synthesis algorithms, and investigate the utility of combining a simple, passive mechanism with an active force control scheme to obtain a compliant manipulator with desirable characteristics for interaction control.