Adaptive Robust Motion Control of Series Elastic Actuator with Unmatched Uncertainties

Yinjie Lin,Zheng Chen,Bin Yao
DOI: https://doi.org/10.1109/ICM46511.2021.9385639
2021-01-01
Abstract:Different from the “stiffer is better” rule in traditional robotic applications, safety and compliance are now drawing more and more attention. Series elastic actuator (SEA), which is an intrinsically safe and compliant actuator, is widely used in robotics. However, compared with the conventional stiff actuators' model, SEA system is high-order nonlinear and simultaneously has unmatched uncertainties owing to the existence of the joint flexibility. These problems make the motion control of SEA more complicated than conventional stiff actuators. To deal with these issues, a precision motion controller is developed in this paper by integrating the adaptive robust control (ARC) and the backstepping design technique. Specifically, the effect of matched and unmatched model uncertainties can be attenuated by the robust law via backstepping design techniques; furthermore, the on-line adaptation law is employed to suppress the parametric uncertainties and further improve the system performance. Theoretically, the tracking performance and stability of the controller are guaranteed. Comparative experiments have been conducted on a designed SEA testbed, and the experimental results validate the effectiveness of the proposed precision motion controller.
What problem does this paper attempt to address?