Title:Construction of Differential-Cascaded Structures for Control of Robot Manipulators

Abstract:This paper focuses on the construction of differential-cascaded structures for control of nonlinear robot manipulators subjected to disturbances and unavailability of partial information of the desired trajectory. The proposed differential-cascaded structures rely on infinite differential series to handle the robustness with respect to time-varying disturbances and the partial knowledge of the desired trajectories for nonlinear robot manipulators. The long-standing problem of reliable adaptation in the presence of sustaining disturbances is solved by the proposed forwardstepping control with forwardstepping adaptation, and stacked reference dynamics yielding adaptive differential-cascaded structures have been proposed to facilitate the forwardstepping adaptation to both the uncertainty of robot dynamics and that of the frequencies of disturbances. A distinctive point of the proposed differential-cascaded approach is that the reference dynamics for design and analysis involve high-order quantities, but via degree-reduction implementation of the reference dynamics, the control typically involves only the low-order quantities, thus facilitating its applications to control of most physical systems. Our result relies on neither the explicit estimation of the disturbances or derivative and second derivative of the desired position nor the solutions to linear/nonlinear regulator equations, and the employed essential element is a differential-cascaded structure governing robot dynamics.