Ellipsoidal constrained state estimation in presence of bounded disturbances

Yasmina Becis-Aubry
DOI: https://doi.org/10.48550/arXiv.2012.03267
2021-04-28
Abstract:This contribution proposes a recursive, computationally efficient, ready-to-use, online method for the ellipsoidal state characterization for linear discrete-time models with additive unknown disturbances vectors (bounded by known possibly degenerate zonotopes) corrupting both the state difference equation and the sporadic measurement vectors, which are expressed as linear inequality and equality constraints on the state vector. The algorithm is decomposed into time updating and observation updating steps. In the latter, a suitable switching estimation gain is designed in such a way as to ensure the input-to-state stability of the estimation error.
Systems and Control
What problem does this paper attempt to address?