Adaptive finite-time backstepping control for a two-wheeled mobile manipulator

This study presents a dynamic modeling and adaptive finite-time backstepping control (AFBSC) strategy for a two-wheeled mobile platform with a three-link manipulator. The Euler-Lagrange method, partially combined with the Newton method, produces a simplified dynamic model wherein the complex coordination transformation process used in traditional mobile platform modeling processes is not required to be considered. Thus, the simplified two-wheeled mobile manipulator achieves fast locomotion and flexible manipulation in the given workspace. Finite-time backstepping virtual errors are introduced into the recursive design procedures to guarantee the rapid convergence of control performance. Furthermore, the uncertainties of coupled nonlinear dynamics are compensated by an adaptive error compensator. Comparative simulations and experiments with an adaptive finite-time sliding mode control (AFSMC) demonstrate the effectiveness of the proposed control scheme. The settling time and variance of tracking error are selected as the criteria that can reflect convergence speed and stability, respectively, under AFBSC and AFSMC, to analyze the experimental data.