Recently various trajectory tracking control methods using acceleration feedback, such as disturbance observer or time delay control, have been proposed for robot manipulators with unknown parameters or disturbances. However, it is not clear why these control methods are robust or what is the essential role of acceleration feedback. In addition, it is important to know what are the problems in these methods.
Our purpose of this paper is to clarify the merit of acceleration information in the case that the signal is available. In the first part of this paper, we make the essential feature of these methods clear, and show what their problems are. In the second part, a robust control scheme using acceleration information is proposed for robot manipulators, which overcomes the above problems. In the end, we discuss the advantage of acceleration information for the proposed scheme.