Fuzzy adaptive control of multivariable nonlinear systems

Abstract

Based on Takagi-Sugeno (TS) fuzzy systems, we present a direct fuzzy model-following adaptive control for multivariable (MIMO) nonlinear systems. The use of the TS fuzzy systems allows the inclusion of a priori information in terms of qualitative knowledge about the plant operating points or analytical conventional linear regulators. It is proven, using Lyapunov stability, that this adaptive scheme is robust against external disturbance, approximation error and input gain variation, and achieves asymptotic tracking of a stable reference model. The effectiveness of the proposed fuzzy approach is demonstrated, by simulation, on a two-link robot model.