Kalman filtering for TS fuzzy state estimation

This paper studies the T-S fuzzy model-based state estimator which the dynamic system can be approximated as linear system. It is suggested for a steady state estimator using standard Kalman filter theory. In that case, the steady state of nonlinear system can be represented by the T-S fuzzy model structure, which is further rearranged to give a set of a linear model. The steady state solutions can be found for a liner model method and dynamic system can be approximated as locally linear system. And then, linear modeled filter is corrected by the fuzzy gain which is a fuzzy system using the relation between the filter residual and its variation. It reduces the measurement residual with noise. Finally, the proposed state estimator is demonstrated on a truck-trailer.