Design of Minimum Variance Fault Diagnosis Filter for Linear Discrete-Time Stochastic Systems with Unknown Inputs

미지입력이 존재하는 선형 이산 활률 시스템의 최소 분산 고장 진단 필터의 설계

In this paper a state reconstruction filter for linear discrete-time stochastic systems with unknown inputs and noises is presented. The suggested filter can estimate the system state vector and the unknown inputs simultaneously As an extension of the filter a fault diagnosis filter for linear discrete-time stochastic systems with unknown inputs and noises is presented for each filters the optimal gain determination methods which minimize the variance of the state reconstruction errorare presented. Finally the usability of the filtersis shown via numerical examples.