MODEL ORDER REDUCTION BY SELECTIVE SENSITIVITY

Many industrial structures are represented by models with a large numb er of degrees of freedom, thus making their use complex and costly. Mo del order reduction alleviates this problem by elaborating lower-dimen sional models that satisfy some properties of the refined model. We pr esent a model order reduction procedure based on the concept of select ive sensitivity. This reduction method has the property of preserving energies and static or dynamic behaviors of the refined model in the r educed model while preserving the essence of its topology.