Vermiglio Simona, Champaney Victor, Sancarlos Abel, Daim Fatima, Kedzia Jean Claude, Duval Jean Louis, Diez Pedro, Chinesta Francisco
ESI Group, 3bis rue Saarinen, 94528 Rungis, France.
PIMM Lab & ESI Group Chair, Arts et Metiers Institute of Technology, 155 Boulevard de Hopital, 75013 Paris, France.
Sensors (Basel). 2020 Oct 5;20(19):5686. doi: 10.3390/s20195686.
Efficient and optimal design of radar-based Advanced Driver Assistant Systems (ADAS) needs the evaluation of many different electromagnetic solutions for evaluating the impact of the radome on the electromagnetic wave propagation. Because of the very high frequency at which these devices operate, with the associated extremely small wavelength, very fine meshes are needed to accurately discretize the electromagnetic equations. Thus, the computational cost of each numerical solution for a given choice of the design or operation parameters, is high (CPU time consuming and needing significant computational resources) compromising the efficiency of standard optimization algorithms. In order to alleviate the just referred difficulties the present paper proposes an approach based on the use of reduced order modeling, in particular the construction of a parametric solution by employing a non-intrusive formulation of the Proper Generalized Decomposition, combined with a powerful phase-angle unwrapping strategy for accurately addressing the electric and magnetic fields interpolation, contributing to improve the design, the calibration and the operational use of those systems.
基于雷达的先进驾驶辅助系统(ADAS)的高效与优化设计需要评估许多不同的电磁解决方案,以评估天线罩对电磁波传播的影响。由于这些设备工作在非常高的频率,且波长极短,因此需要非常精细的网格来精确离散化电磁方程。因此,对于给定的设计或操作参数选择,每个数值解的计算成本都很高(耗费CPU时间且需要大量计算资源),这影响了标准优化算法的效率。为了缓解上述困难,本文提出了一种基于降阶建模的方法,特别是通过采用适当广义分解的非侵入式公式构建参数化解,并结合一种强大的相角展开策略来精确处理电场和磁场插值,有助于改进这些系统的设计、校准和操作使用。