A novel source reconstruction method from near-field based on DE-PSO algorithm

In the field of electromagnetics, near-field scanning was initially applied to antenna near-to-far-field transformations and antenna array element testing. Over years of research, near-field scanning has developed mature theoretical frameworks and practical expertise in the antenna domain. In recent years, near-field scanning and radiation source reconstruction techniques have increasingly been applied to PCB-level Electromagnetic Interference analysis. This is because EMC test engineers could not access the complete design schematics of the circuit boards of the DUT. To address challenges such as far-field prediction and EMI port coupling power calculations, after obtaining electromagnetic near-field scanning results of a PCB, engineers typically perform dipole reconstruction of radiation sources on the board. The reconstructed equivalent dipole models are then utilized for subsequent predictive analyses. Traditionally, solution methods such as least squares and singular value decomposition have been employed for dipole reconstruction of radiation sources using planar sampling data. However, since the least-squares method is suitable for linear models while dipole reconstruction inherently involves nonlinear modeling, and the objective function of dipole reconstruction problems often exhibits multimodality, the least-squares approach tends to produce significant errors. This paper proposes a novel algorithm that uses the DE-PSO hybrid algorithm for dipole reconstruction, which balances global search capability with rapid convergence. The method has shown promising results in simulation data.