Enhancement of Vibration Suppression for Multi-Link Flexible Robots Using Direct Parameter Adaptation

The precise modeling of multi-link flexible robot is hard to obtain and even if we can get it, it is difficult to use in on-line control tasks. Due to these reasons, a simplified model is often used to describe the robot dynamics instead of using complex dynamic model. In the simplified model, however, parameter mismatch and structural modeling error are inevitable. To compensate these uncertainties, an adaptive control is formulated in this paper after separating the system into rigid and flexible subsystems. A simple direct parameter update rule is presented considering the flexible subsystem. Different from most of adaptive control schemes in multi-link flexible robots, which is indirect model-independent approach, the proposed adaptive control is a direct one and good for fast suppression of vibration of uncertain and untuned systems. We verify the effectiveness of the proposed algorithm through experiments.