Estimador de parámetros para sistemas de orden superior

Abstract

This paper~presents parameter estimations considering~an observable~black box~system signal represented with high order finite differences. The model is~expressed as a~sequence of delayed observable states~conforming space state vector~to its~known associated parameters. The estimation technique is based on~pseudo-inverse~and an innovation process~associated with perturbation and an observable signal. The simulation~uses two high order~finite difference~models, second and third grades with associate parameters. In both cases the observed signal~depends~on their delayed states. Therefore, it is necessary to estimate two~and three parameters with respect to second and third order models, respectively. The estimation results allowed identifying~the observable signal with a high convergence rate in base of functional error, as~shown in~the figures. Instrumental~variable~is the technique applied to the second~probability~moment constructing~the estimator~for a~high finite difference~order system.