Robust and non-asymptotic state estimation for MIMO descriptor systems.

Abstract

In this research paper, a state estimation framework for a class of descriptor linear systems with MIMO is provided by using auxiliary modulating dynamical systems. First, the considered model is transformed into a simpler form involving the derivatives of inputs and outputs, based on which the auxiliary systems are applied. Then, the state variables are expressed through modulating integrals without the need for initial conditions, guaranteeing non-asymptotic convergence within fixed-time. This framework does not require the calculation of the derivatives of noisy outputs in discrete cases, reducing sensitivity to high-frequency noise in the estimation. Finally, the performance of the proposed method is validated through numerical simulations, which provide practical insights into its effectiveness and enable a comparison with some observers.