Composite Hierarchical Anti-Disturbance Control for Robotic Systems with Multiple Disturbances

Various sources of disturbances exist simultaneously in robotic systems, such as vibrations, frictions, measurement noises, and equivalent disturbances from unmodeled dynamics and nonlinearities. However, most results on anti-disturbance control focus on only one type of disturbances, which cannot reflect the real applications and may lead to design conservativeness due to partial use of the disturbance information. In this paper, we propose a composite hierarchical anti-disturbance control (CHADC) strategy for robotic systems in the presence of multiple disturbances as well as system uncertainties. Particularly, we assume the existence of two types of disturbances, where the first type represents disturbances from exogenous systems with model perturbations, while the second type includes other random disturbances satisfying the L2-norm bound condition. Accordingly, the CHADC control architecture is composed of a nonlinear disturbance observer (NDO) and an H-infinity based PID controller, where the NDO is constructed to estimate the first type of disturbances and provide feed forward compensation, while the feedback PID loop is optimized using H-infinity theory to minimize the second type of disturbances. Robustness against system uncertainties is also considered in this hierarchical control structure. The proposed control approach is applied to a two-link robotic manipulator and compared with the conventional DOBC (disturbance observer based control) strategies.