Robustness of Online Adaptive Control Framework Design for a Multi-Fingered Robotic Hand Presence of Nonlinearities and Disturbances

Abstract

This paper presents the development and evaluation of an online robust adaptive control framework for a 14-degree-of-freedom (DOF) multi-fingered robotic hand (MFRH). Although standard adaptive control can provide asymptotic tracking under ideal disturbance-free conditions, persistent disturbances and modeling uncertainties may cause parameter drift, which can degrade tracking performance and compromise closed-loop robustness. To address this issue, the proposed framework is developed based on the nonlinear Lagrangian dynamics of the 14-DOF MFRH and incorporates a robust adaptation law using the $\sigma $ -modification technique to limit parameter drift under the considered disturbance conditions. A Lyapunov-based stability analysis establishes that both the tracking errors and parameter estimates are uniformly ultimately bounded (UUB) under the stated assumptions. The proposed controller is evaluated through comparative simulations against a standard adaptive controller under multi-frequency sinusoidal tracking and pulse-disturbance conditions. The results demonstrate that the $\sigma $ -modification limits parameter drift and maintains bounded tracking behavior under the considered disturbance conditions, while also revealing a trade-off between nominal tracking accuracy and robustness. These results demonstrate the applicability of the proposed robust adaptive framework to the considered multi-fingered robotic-hand simulation conditions.