University of Delaware - College of Engineering
ELECTRICAL & COMPUTER ENGINEERING

Research

Photonics & Electromagnetics

Evolutionary optimization of electromagnetic devices

Daniel Weile

The design of electromagnetic devices is a time-consuming and arduous task. Though computational electromagnetics techniques have eliminated some of the burden of laboratory work, in many cases they have merely moved the exertion from the anechoic chamber to the computer room. The research done under this project seeks to rectify this situation by letting computers do the design directly using optimization paradigms inspired by nature.

Foremost among these is Darwinian evolution, nature's method for maximizing survivability. Optimizers based on evolution are known as evolutionary optimizers. While they vary greatly in their details, evolutionary optimizers all work by considering potential designs of some object as an individual in nature, and basing the survival of the individual on its performance in solving the problem at hand. New designs are created through both mutation and hybridization, just as in nature.

Other nature-based optimization paradigms include simulated annealing (which is based on the cooling of metals), particle swarm optimization (which is based on social interactions), and ant colony optimization (which is based on the method ants use to find food). The application of these methods to electromagnetic optimization problems has resulted in marvelous new electromagnetic devices ranging from antennas and absorbers to radomes and photonic crystals.

Recent publications

S. Cui, D. S. Weile, and J. L. Volakis "Novel planar electromagnetic absorber design using genetic algorithms," IEEE Trans. Antennas Propagat., vol. 54, no. 6, pp. 1811-1817, 2006.

S. Cui and D. S. Weile, "Application of a particle swarm optimization scheme to the design of electromagnetic absorbers," IEEE Trans. Antennas Propagat., vol. 53, no. 11, pp. 3616-3624, 2005.

S. Cui, A. Mohan, and D. S. Weile, "Pareto optimal design of absorbers using a parallel elitist nondominated sorting genetic algorithm and the finite element boundary integral method" IEEE Trans. Antennas Propagat., vol. 53, no. 6, pp. 2099-2107, 2005.


Bookmark and Share