Polynomial Chaos-Based Macromodeling of Multiport Systems using an Input-Output Approach

An innovative technique to build stochastic frequency-domain macromodels of generic linear multiport systems is presented. The proposed method calculates a macromodel of the system transfer function including its statistical properties, making it tailored for variability analysis. The combination of the modeling power of the Vector Fitting algorithm with the Polynomial Chaos expansion applied at an input–output level allows to accurately and efficiently describe the system variability features. Thanks to its versatility and automated order selection, the proposed technique is suitable to be applied to a large range of complex modern electrical systems (e.g., filters and interconnections) and can tackle the case of correlated random variables. The performance in terms of accuracy and computational efficiency of the proposed method is compared with respect to the standard Monte Carlo analysis for two pertinent numerical examples.