Block Structure Preserving Model Order Reduction for A-EFIE Integral Equation Method
🤖 AI Summary
This work addresses the degradation of accuracy and subspace redundancy arising from the destruction of the inherent block structure during model order reduction (MOR) of the augmented electric field integral equation (A-EFIE). To preserve the original 2×2 block structure, we propose a structure-preserving MOR method that constructs dedicated, coupled constraint subspaces—separately for the electric current and auxiliary variables—enabling projection-based reduction without structural disruption. Unlike conventional global subspace approaches, our method achieves significant model-size compression (over 60% reduction observed empirically) while simultaneously improving broadband response accuracy and numerical stability. Numerical experiments on high-frequency electromagnetic scattering problems demonstrate the method’s efficiency and robustness. The proposed framework establishes a scalable, structured MOR paradigm for fast A-EFIE-based electromagnetic simulations.
Application Category
📝 Abstract
A Block Structure Preserving Model Order Reduction approach is proposed for Integral Equations methods based on the Augmented Electric Field Integral Equation. This approach allows for representing the unknown fields with dedicated subspaces. Numerical results show that this leads to smaller reduced-order models and higher accuracy.
Problem
Research questions and friction points this paper is trying to address.
Preserving block structure in model order reduction
Improving accuracy of integral equation methods
Reducing computational complexity for electromagnetic simulations
Innovation
Methods, ideas, or system contributions that make the work stand out.
Block structure preserving model order reduction
Augmented electric field integral equation method
Dedicated subspaces for unknown field representation
🔎 Similar Papers
No similar papers found.
