On state and parameter estimation via innovation representation with applications to aquatic ecosystems

Abstract

The problem of simultaneous state and system parameter estimation for aquatic ecosystems is considered in this paper. This problem has been solved through the identification of an innovation dynamical model representation of the underlying process within the hydrological cycle of the ecosystem basin. A recursive prediction error (RPE) algorithm is derived for the joint system parameter and state estimation through minimization of the innovation variance (MIV). On the grounds of the case study on Ohridean Lake using recorded data by the Hydro-meteorological Institution a comparison analysis has been carries out between the methods of innovation variance minimization and of extended non-linear Kalman filter.