Fast Yield Estimation of Multi-Band Patch Antennas by PC-Kriging
Xiaosong Du, Leifur Þ. Leifsson, S. Koziel
May 1, 2019
Citations
3
Citations
Journal
2019 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization (NEMO)
Full text
Semantic Scholar
Key Takeaway Key Takeaway: The PC-Kriging method accurately estimates multi-band patch antenna yields with fewer training points, resulting in significant computational savings without compromising evaluation reliability.
Abstract
The PC-Kriging metamodeling method is proposed for yield estimation of multi-band patch antennas. PC-Kriging is a combination of polynomial chaos expansion (PCE) and Kriging metamodeling, where PCE is used as a trend function for the Kriging interpolation metamodel. The method is demonstrated on the Ishigami analytical function and a dual-band patch antenna. The PC-Kriging is shown to reach the prescribed accuracy limit with significantly fewer training points than both PCE and Kriging. This translates into considerable computational savings of yield estimation over alternative metamodel-based procedures and direct EM-driven Monte Carlo simulation. The saving are obtained without compromising evaluation reliability.