Self-Organized Neural Network-Based Control for Uncertain Electro-Hydraulic Servo Actuator With Output Constraint and Input Saturation

To significantly improve the tracking performance of electro-hydraulic servo actuators (EHSAs) subject to unknown disturbances, uncertainties, and unstructured system dynamics, this paper proposes a control strategy based on a self-organizing radial basis function neural network (SORBFNN) augmented by a finite-time disturbance observer (DO). The SORBFNN, which employs time-varying Gaussian basis functions, is first utilized to effectively approximate unstructured control coefficients. Simultaneously, a DO is designed to suppress external disturbances and mismatched uncertainties. Furthermore, to further enhance system robustness, output constraints and input saturation techniques are incorporated into the design. The stability of the closed-loop system is rigorously demonstrated using the Lyapunov stability theorem, and the method’s effectiveness is experimentally validated through comparative benchmarks on a rotary EHSA.