Design optimization techniques for printed antennas and periodic structures.

Abstract

This thesis demonstrates a comprehensive study of optimization algorithms combined with finite element-boundary integral simulations on microstrip antennas and periodic structures. Among the various optimization methods considered, the sequential quadratic programming (a class of gradient-based algorithms) is applied to several size optimization problems; and two gradient-free methods (genetic algorithms and simulated annealing) are employed for shape and topology optimization. Recent developments in fast algorithms now offer the possibility of designs and moreover allow for full flexibility in material specification across three dimensions. The first part of the thesis considers various antenna designs already investigated but not optimized before. The examples include ferrite antennas, stacked patches, a patch with a folded-slot feed, and irregular-shaped broadband/dual-band patches. These designs serve to validate the optimization process and develop confidence prior to proceeding with more complex design problems. The second part of the thesis deals with the design of novel antennas and periodic structures. Frequency selective surface (FSS) elements are designed for a high-pass filter application, as well as a flat-phase element which is demonstrated to achieve much higher bandwidth for a dipole array placed on top of it. Finally, shape and material optimization are combined to design a full 3-D photonic bandgap (PBG) structure which substantially increases the bandwidth of a patch residing on the PBG substrate.