An estimator of the inverse covariance matrix and its application to ML parameter estimation in dynamical systems (Q1592898)

The paper deals with parameter estimation in a dynamical system. The system under consideration is described by a phenomenological model which is based on the prior knowledge of the physical phenomena that are supposed to take place in the system. The model is written under the form of a differential parametric deterministic state-space representation.The parameter estimation problem is to estimate the parameter values from input and state data in the presence of additive correlated noise on the state measurements. For this purpose the authors obtain an exact formula of the inverse covariance matrix (ICM) of an autoregressive stochastic process, using the Gohberg-Semencul explicit inverse of the Toeplitz matrix. This formula is used to build an estimator of the inverse covariance matrix of a stochastic process based on a single realization. This estimator is applied to maximum likelihood parameter estimation in the dynamical system under consideration. It is demonstrated that when both the deterministic and the stochastic parts of the system belong to the model set, the statistical properties of the estimated parameters converge to the Cramér-Rao lower bound. The algorithm is also valid if uncertainties cannot be exactly described by an autoregressive process.