Abstract: An approach of self-learning fuzzy sliding-mode control , which uses the fuzzy logic systems to approximate sliding-mode equivalent control, is presented for the tracking control of a class of systems with the nonlinear and parameter uncertainties. The fuzzy control rules are updated through on-line learning, which make the output of fuzzy control approach sliding-mode equivalent along the direction of making sliding-mode asymptotic stable. The asymptotic stability of self-learning fuzzy sliding-mode control systems is proved based on Lyapunov theory. The proposed method is applied to an electrohydraulic servo systems, the results of simulation and experiment of real-time control show that the proposed method is distinctly excelled than that of conventional sliding-mode control and satisfied control precision and stability can be obtained.

Key words: Electrohydraulic servo systems, Self-learning fuzzy sliding-mode control, Simulation and experiment