Fuzzy state estimation in linear dynamic systems

Abstract

The problem of state estimation in linear dynamic systems with fuzzy initial states and poorly defined perturbation input signals (i.e fuzzy inputs) is considered. A new notion of vector summ of n-dimensional fuzzy sets provides the basis for state and input "imprecision" propagation trough the systems dynamics. Recursive state estimation formulae are developed to obtain a fuzzy set of possible states at each instant of time.