New Adaptive Multi-Expansion Frequencies Approach for SP-MORe Techniques With Application to the Well-Conditioned Asymptotic Waveform Evaluation

Abstract

Fast frequency sweep using model order reduction (MORe) techniques is one of the most valuable and efficient features in high-frequency simulators. These techniques usually use a single full solution at a given frequency to build a smaller approximation reduced model, which might be insufficient for wideband and ultrawideband simulations of high quality factor RF and microwave devices. Despite the fact that many computationally reliable error estimate methods have been presented, they all share the same starting point, which is to estimate the residual error between fast sweep solutions and not the actual error. In this paper, we introduce a new accurate and computationally reliable error estimate approach based on the lossless network condition for the automation of single-point MORe. When applied with the well-conditioned asymptotic waveform evaluation, the new approach shows very good performances in terms of accuracy and computation time compared with existing multipoint MORe and commercial EM software.