Robust identification of linear state-space models in presence of component and sensor faults

Abstract

Joint estimation of states and time-varying parameters of linear state space models is of practical importance for fault diagnosis and fault tolerant control. This paper considers robust identification of linear state-space models with component and sensor faults. On the other side, previous works on this topic have not considered joint estimation of linear systems in presence of outliers. They can significantly make worse the properties of linearly recursive algorithms, which are designed to work in the presence of Gaussian noises. Because of their good features in robust filtering, the modified Masreliez-Martin filter represents a cornerstone for realization of the robust algorithm for joint state-parameter estimation of linear time-varying stochastic systems in presence of non-Gaussian noises. The good features of the proposed robust algorithm for joint estimation of linear time-varying stochastic systems is illustrated by simulations.