Robust Control of Electro-Hydraulic Systems Subject to Input Constraints

This work presents a robust, neural network based controller formulation for the displacement tracking problem of electro-hydraulic systems subject to uncertainties associated with their dynamical parameters and actuator saturation. Specifically, a neural network based compensator is utilized to estimate some of the nonlinear components of the uncertain dynamical terms and then in conjunction with robust backstepping procedure, the overall formulation ensure the uniform practical stability of the closed-loop system. Stability and convergence of error terms are proven using Lyapunov based arguments and numerical studies are presented in order to illustrate the feasibility of the proposed methodology.