Skip to Main Content
In recent years, evolutionary optimization methods such as genetic algorithms have become increasingly popular in EM optimization problems. However, these methods tend to have slow convergence bounds and further, they do not yield a deterministic optimal solution. In this paper, we propose a new method of using Kriging meta-modeling in conjunction with the divided rectangles (DIRECT) global optimizer to yield a global optimum solution. DIRECT yields a deterministic answer with a fast convergence bound and inherits both local and global optimization properties. This proposed hybrid optimization routine is applied here to two examples dealing with electromagnetic coupling reduction. One of them deals with minimizing coupling to sensors from antenna radiation on automobiles. The other example focuses on multi-sensor optimization subject to RF coupling constraints.