Smoothed estimation of unknown inputs and states in dynamic systems with application to oceanic flow field reconstruction
Huazhen Fang, Raymond A. DE CALLAFON, Peter J. S. Franks · International Journal of Adaptive Control and Signal Processing · 2014
Summary Forward‐backward smoothing based unknown input and state estimation for dynamic systems is studied in this paper, motivated by reconstruction of an oceanographic flow field using a swarm of buoyancy‐controlled drifters. The development is conducted in a Bayesian framework. A Bayesian paradigm is constructed first to offer a probabilistic view of the unknown quantities given the measurements. Then a maximuma posterioriis established to build a means for simultaneous input and state smoothing, which can be solved by the classical Gauss–Newton method in the nonlinear case. Application to reconstruction of a complex three‐dimensional flow field is presented and investigatedviasimulation studies. Copyright © 2014 John Wiley & Sons, Ltd.