Kalman 필터 기반의 퍼지 적응상태 추정 기법

A Fuzzy State Estimation Technique Based on Kalman Filter

Kalman filter algorithm has widely been used under the assumption that a system noise exerted on a stochastic system is white Gaussian. However, the Kalman filter is not converged on the true state but eventually diverged when an unknown abrupt input is applied to the system. In order to solve filter divergence problem, in this paper, a fuzzy state estimation technique compensating filtered estimates corresponding to the abrupt unknown input by using innovation process is introduced. The suggested estimation technique is guaranteed through computer simulation experiments.