Nonlinear disturbance observer–based adaptive neural control for electro-hydraulic servo system with model uncertainty and full-state constraints

The electro-hydraulic servo system (EHSS) performs model uncertainty and state constraints such that the exact model-based controller is difficult to be designed. In this work, a nonlinear disturbance observer (NDO)-based adaptive neural control (ANC) is proposed for the EHSS, in which a nonlinear transformation function is constructed to make the state constraints problem transformed into state unconstraint problem. The NDO is introduced to improve the disturbance rejection ability. The ANC is utilized to approximate unmodeled dynamics. The second-order filters are integrated with backstepping control to solve the explosion of complexity. The proposed NDO-based ANC scheme confines all states within the predefined bounds, improves the robustness of closed-loop system, and alleviates the computation burden. Moreover, the stability analysis for the closed-loop system is given within the Lyapunov framework. Simulations and experiments show that the proposed control scheme can achieve excellent control performance and robustness with regard to full-state constraints and model uncertainty.