Robust fault estimation of nonlinear systems using SOS approach

This paper proposes a new method of robust fault estimation (FE) for a class of nonlinear systems represented by polynomial fuzzy models. Existing results on these systems essentially focus on fault detection and isolation (FDI) approach. The sufficient conditions for designing gains of the estimator are established via Lyapunov theory, which are formulated in terms of polynomial matrix inequalities (PLMIs). These PLMIs are solved using sum of squares (SOS) approach and estimator gains are calculated using SOSTOOLS toolbox in Matlab. The effectiveness of the proposed approach has been illustrated considering a numerical example and has been found to be satisfactory.