Inescapable-set Estimation for Nonlinear Systems with Non-vanishing Disturbances

A common approach for stabilization and performance objectives in control design for TS/polynomial fuzzy models is to solve an LMI problem which proves global stability and an upper bound of a performance function for any shape of membership functions. However, fuzzy models are local and, obviously, a larger modelling region yields lower performance bounds. In systems subject to disturbances, determining the optimal size of the modelling region is needed in order to avoid small ones (useless, as state escapes from them) or overly large ones (suboptimal performance). This paper addresses such problem and suggests a multicriteria fuzzy controller strategy design in order to ensure disturbance rejection performance without leaving a particular modelling region.