In this paper, a procedure for the recursive approximation of the feas
ible parameter set of a linear model with a set membership uncertainty
description is provided. Approximating regions of parallelotopic shap
e are considered. The new contribution of this paper consists in devis
ing a general procedure allowing for block professing of q > 1 measure
ments at each recursion step. Based on this, several approximation str
ategies for polytopes are presented. Simulation experiments are perfor
med, showing the effectiveness of the algorithm as compared to the ori
ginal algorithm processing one measurement at each step. (C) 1998 Else
vier Science Ltd. All rights reserved.