Fast-convergence nonlinear observer-based neural adaptive robust control for hydraulic servo position system

Parameter uncertainty and uncertain nonlinearity seriously influence the trajectory tracking control performance of hydraulic servo systems. To address this challenge, a fast-convergence nonlinear observer-based neural adaptive robust control strategy is proposed. Firstly, a faln nonlinear function-based fast-convergence observer is designed in this paper. The observer has two key advantages: (1) fast convergence rate of the closed-loop errors on a global scope, (2) high estimation accuracy. Especially, to deal with uncertain disturbances, an adaptive neural network and observer are used to estimate in parallel, which reduces the estimation burden of the observer. Additionally, a neural adaptive robust controller is proposed to solve the physical parameter uncertainty of the system and suppress (compensate) the unexpected nonlinear disturbance. More importantly, the stability of the whole closed-loop system is guaranteed by Lyapunov theory. Finally, the effectiveness of the proposed nonlinear observer is verified by numerical simulation. Comparative motion experiments were carried out under low-frequency, high-frequency, additional external disturbance and triangular wave. The experimental results show that the proposed strategy significantly improves the trajectory tracking performance and robustness of hydraulic servo system.