Robustness of Smooth Sliding Control to Unmodeled Dynamics: Overcoming First-Order SMC and Super-Twisting Algorithm

Abstract

In this paper we revisit the earlier smooth sliding control (SSC) by presenting the design and an alternative approach to analyze its robustness with respect to unmodeled dynamics. The control approach relies on some appropriate prediction error and a filtered control signal to obtain a smooth output-feedback based on first-order sliding mode control law for chattering avoidance. Global exponentially stability with respect to a small compact residual set, practical trajectory tracking and ideal sliding modes are guaranteed by means of a small-gain analysis, even in the presence of adverse parasitic dynamics. Comparison results discuss to what extent the modified SSC architecture may serve as an alternative to classical first-order sliding modes (FOSM) and super-twisting algorithm (STA). In fact, simulations show that SSC outperforms FOSM and STA for the same numerical examples recently used in the lively debate between continuous and discontinuous sliding mode control options for chattering alleviation. The main purpose here is to stir new research about robustness improvements in both classical and on higher-order sliding mode algorithms.