Remarks on validation of uncertainty models using frequencyresponse data

Uncertainty validation using frequency response data has been studied by several authors. If the uncertainty is assumed to be stable, then the problem amounts to one of boundary interpolation. It is shown in this paper that for more general uncertainty models, an additional constraint is needed in order for the results of Boulet et al. (1998) and Chen (1997) to be applicable. Boundary interpolation, together with the proposed constraint, gives a satisfactory answer to the validation problem for uncertainty models