A model order reduction technique for parametric uncertain models

Applying model-order reduction algorithms developed for fixed-parameter models based-on balancing transformations to parametric uncertain models is considered in this paper. High-order models are first decomposed into sub-models of lower order using serial and parallel decompositions, then model-order reduction algorithms implemented in a symbolic environment are used to reduce these sub-models. At the end of this process, a reduced-order model of the original higher-order model is produced by recombining the reduced sub-models.