Optimal finite-dimensional recursive identification in a polynomial output mapping class (Q794579)

The search for optimal finite-dimensional recursive filters in nonlinear differential systems was successful in some restricted system classes [e.g. refer to \textit{S. Marcus} and \textit{A. Willsky}, SIAM J. Math. Anal. 9, 312-327 (1978; Zbl 0377.93056)]. This paper considers a new system class admitting an optimal finite dimensional recursive estimator. An explicit filter structure is constructed and two examples are discussed to show the application. A time-varying polynomial output equation expressing an implicit dependence of the observed signal on the parameter vector admits a finite-dimensional recursive representation for the optimal on-line estimator. Both continuous and discrete observations result in differential systems which give the estimate. An exponentially weighted least-squares formulation is used. Exponential stability of the identifier and convergence to the true parameter value are proved. A time-continuous algorithm for real time identification of two dissociation constants of a pH-control process is presented. The discrete identification is applied in a dynamic ARMA-type difference model which is quadratic in the parameter to be estimated.