Stable Gain Adaptation for Time-Delay Control of Robot Manipulators

Abstract

This paper proposes adaptive gain dynamics for time-delay control (TDC) of robot manipulators. The TDC is a widely employed approach for control of robot manipulators because it is model-independent, simple, and robust. Recently, however, it is reported that TDC with excessively high gains causes unstable or oscillated responses. To overcome this limitation, this study designs adaptive gain dynamics for the TDC using a sliding variable and an acceptance layer. When the gain dynamics automatically adjusts a control gain according to system states, the adaptive control gain moves into a stable range. Therefore, the TDC with the proposed gain dynamics becomes adaptive and stable. The effectiveness of the proposed gain dynamics has been verified by simulations and experiments.