Neurofuzzy state identification using prefiltering

Download

Postscript - Pre print Download (332Kb)

Description/Abstract

A new state estimator algorithm is introduced based on a neurofuzzy network and the Kalman filter algorithm. The major contribution of the paper is recognition of a bias problem in the parameter estimation of the state space model and the introduction of a simple ffective pre-filtering method to achieve unbiased parameter estimates in the state space model which will then be applied for state estimation using the Kalman filtering algorithm. Fundamental to this method is a simple pre-filtering procedure using a non-linear principal component analysis PCAmethod based on the neuro-fuzzy basis set. This preltering procedure can be performed withoutprior system structure information. Some numerical examples are included to demonstrate theeffectiveness of the new approach.