An adaptive dynamic inversion-extremum seeking control approach for constrained robotic motion tasks

In this paper, an adaptive control approach is proposed for performance of constrained robot end-effector movements in presence of uncertainty. In real-world scenarios, complex physical phenomena occuring at the place of interaction may introduce nonlinearities in the system dynamics, which have to be taken into account for proper system control. We currently propose an Extremum Seeking (ES) Model Reference Adaptive Control (MRAC) approach for state tracking of multiple-input multiple-output systems which enclose nonlinearities in their dynamics and involve parametric uncertainty by employing Adaptive Dynamic Inversion (ADI). According to ADI, system nonlinearities are assumed known and are taken into account in the design of the system control law. The proposed scheme is based on MRAC and ADI while the unknown controller parameters are adapted by ES control. The system is shown to achieve global and asymptotic reference state tracking under the proposed control law by performing Lyapunov and averaging analysis. The approach is evaluated in simulation and in an experimental robot task.