Fuzzy adaptive control enhancement for non-affine systems with unknown control gain sign

Abstract

A fuzzy adaptive controller scheme is presented in this work, where the system dynamics are estimated in real time by mean of fuzzy systems. The controller design is dedicated to a class of nonlinear SISO systems with unknown model and control gain sign. Such a control design becomes a true challenge, and generally impossible via the inversion if the system is non-affine. This problem is resolved by using the mean value theory to render the system model affine in the control and introducing a Nussbaum function in order to obtain an estimation of the control gain sign. A simulation example illustrates the performance enhancement realized by the present control strategy.