A fuzzy neural network based extended Kalman filter

Abstract

Abstract In this paper we construct a discrete extended Kalman filter by using fuzzy neural networks. The constructed filter makes it possible to estimate the states of a nonlinear dynamical system with unknown plant model. The unknown plant is identified by a fuzzy neural network and the modelling error can be compensated by a simple but efficient method that avoids the occurrence of divergence in state estimation. A computer simulation is presented to illustrate the performance and applicability of the proposed filter.