Fuzzy Modeling for Nonlinear System with Structured Data Uncertainty

In this paper, a new fuzzy modeling method with structured data uncertainty is proposed for nonlinear systems. The new robust modeling criterion, which aims to minimize the worst-case residual error, is formulated. The algorithm to obtain the approximate solutions of modeling problems under the new criterion is given in consideration of data uncertainty with a known structure. The new method is superior to the existing nonlinear modeling approaches. The advantages of the new method are illustrated by simulation results.