Development and Application of Genetic Programming in the Design and Optimization of Ultra-Wideband 3D Metamaterials

Created by W.Langdon from gp-bibliography.bib Revision:1.4163

@PhdThesis{Rayno:thesis,
  author =       "Jennifer H. Rayno",
  title =        "Development and Application of Genetic Programming in
                 the Design and Optimization of Ultra-Wideband {3D}
                 Metamaterials",
  school =       "University of Hawaii at Manoa",
  year =         "2016",
  address =      "USA",
  month =        may,
  keywords =     "genetic algorithms, genetic programming, 3D
                 metamaterial synthesis",
  URL =          "https://search.proquest.com/docview/1846950904?pq-origsite=gscholar",
  URL =          "http://hdl.handle.net/10125/51406",
  URL =          "https://scholarspace.manoa.hawaii.edu/handle/10125/51406",
  size =         "194 pages",
  abstract =     "Metamaterials are materials with engineered
                 characteristics and unique properties not naturally
                 available, such as artificial magnetic conductors
                 (AMC). Limitation of present AMC designs is related to
                 their narrowband and high frequency operation, in GHz
                 range. For many commercial and military applications,
                 however, it is desired to design such materials in
                 lower MHz band and with ultra-wideband (UWB)
                 performance. In addition, typical 2D AMCs are designed
                 by trial and error, often based on combination of
                 layers of existing designs, and lossy materials are
                 used to achieve broadband performance. There is no
                 methodology that exists for designing true-3D
                 metamaterials with broadband characteristics in the MHz
                 band. This research uses genetic programming (GP) to
                 automatically and efficiently explore the use of 3D
                 design space to develop materials with the desired low
                 frequency and broadband characteristics. Genetic
                 programing is a genetically based evolutionary process
                 that creates and modifies new geometries to achieve
                 final designs that meet desired specifications. In this
                 dissertation, GP software is developed and used to
                 synthesize 3D, compact, UWB AMC ground planes, with a
                 focus on achieving a lower frequency response and
                 without using lossy or expensive magnetic materials.
                 Full-wave electromagnetic simulation software (HFSS) is
                 used to evaluate these designs. To accelerate the
                 design process, GP is hybridized with a low-level
                 optimizer, where GP creates and modifies topologies at
                 the upper level while at the lower level each design is
                 optimized separate from GP. The code is further
                 parallelized to speed up the computations. Simulation
                 results for nine AMC ground plane examples meeting the
                 specifications (225-450 MHz, compact) with thicknesses
                 ranging from lambda0/11 to lambda_0/16 are presented to
                 illustrate variety of successful topologies achieved by
                 GP software while requiring only a set of design
                 specifications. This research thus provides an
                 efficient design methodology for electromagnetic
                 devices and systems, when augmented with suitable
                 design algorithms, it could be used to design 3D
                 metamaterials in general, antennas, and antenna array
                 systems. Results from this research specifically fill a
                 significant need of designing lower frequency UWB AMC
                 ground planes without the use of heavy and/or expensive
                 magnetic materials typically used in the MHz range.",
  notes =        "ProQuest 10295933

                 supervisor Magdy F. Iskander",
}

Genetic Programming entries for Jennifer Taylor Rayno

Citations