Title :
Time-domain H∞ identification
Author :
Theodor, Y. ; Shaked, U. ; Berman, N.
Author_Institution :
Fac. of Eng., Tel Aviv Univ., Israel
fDate :
7/1/1996 12:00:00 AM
Abstract :
The problem of parameter identification, for single-input, single-output ARX systems, is considered. Recent results in H∞-nonlinear filtering are used to formulate a nonlinear H∞ time-domain prediction-error-modeling (PEM) identification method. The performance of the new method is guaranteed by a preassigned bound on the ratio between the energy of the prediction error of the obtained model and the energy of the exogenous disturbances. The potential usefulness of the H∞ time-domain identification method is illustrated by a numerical example
Keywords :
H∞ optimisation; autoregressive processes; discrete time systems; filtering theory; linear systems; nonlinear filters; parameter estimation; H∞-nonlinear filtering; exogenous disturbances; nonlinear H∞ time-domain prediction-error-modeling identification method; parameter identification; prediction error; single-input single-output ARX systems; Automatic control; Error correction; Estimation error; Filters; Noise measurement; Predictive models; Signal processing; System identification; Time domain analysis;
Journal_Title :
Automatic Control, IEEE Transactions on
