published_or_final_version / abstract / toc / Mechanical Engineering / Master / Master of Philosophy

This thesis presents a wide spectrum of novel extensions and enhancements to critical components of modern electromagnetic analysis and design systems. These advancements are achieved through the use of computational intelligence, which comprises neural networks, evolutionary algorithms, and fuzzy systems. These tools have been proven in myriad industrial applications ranging from computer network optimization to heavy machinery control. / The analysis module of an electromagnetic analysis and design system typically comprises mesh generation and mesh improvement stages. A novel method for discovering optimal orderings of mesh improvement operators is proposed and leads to a suite of novel mesh improvement techniques. The new techniques outperform existing methods in both mesh quality improvement and computational cost. / The remaining contributions pertain to the design module. Specifically, a novel space mapping method is proposed, which allows for the optimization of response surface models. The method is able to combine the accuracy of fine models with the speed of coarse models. Optimal results are achieved for a fraction of the cost of the standard optimization approach. / Models built from computational data often do not take into consideration the intrinsic characteristics of the data. A novel model building approach is proposed, which customizes the model to the underlying responses and accelerates searching within the model. The novel approach is able to significantly reduce model error and accelerate optimization. / Automatic design schemes for 2D structures typically preconceive the final design or create an intractable search space. A novel non-preconceived approach is presented, which relies on a new genome structure and genetic operators. The new approach is capable of a threefold performance improvement and improved manufacturability. / Automatic design of 3D wire structures is often based on "in-series" architectures, which limit performance. A novel technique for automatic creative design of 3D wire antennas is proposed. The antenna structures are grown from a starting wire and invalid designs are avoided. The high quality antennas that emerge from this bio-inspired approach could not have been obtained by a human designer and are able to outperform standard designs.

Optimal partitioning of three-dimensional (3-D) mesh applications necessitates dynamically determining and optimizing for the most time-inhibiting factors, such as load imbalance and communication volume. One challenge is to create an analytical model where the programmer can focus on optimizing load imbalance or communication volume to reduce execution time. Another challenge is the best individual performance of a specific mesh refinement demands precise study and the selection of the suitable computation strategy. Very-large-scale finite element method (FEM) applications require sophisticated capabilities for using the underlying parallel computer's resources in the most efficient way. Thus, classifying these requirements in a manner that conforms to the programmer is crucial. / This thesis contributes a simulation-based approach for the algorithm analysis and design of parallel, 3-D FEM mesh refinement that utilizes Petri Nets (PN) as the modeling and simulation tool. PN models are implemented based on detailed software prototypes and system architectures, which imitate the behaviour of the parallel meshing process. Subsequently, estimates for performance measures are derived from discrete event simulations. New communication strategies are contributed in the thesis for parallel mesh refinement that pipeline the computation and communication time by means of the workload prediction approach and task breaking point approach. To examine the performance of these new designs, PN models are created for modeling and simulating each of them and their efficiencies are justified by the simulation results. Also based on the PN modeling approach, the performance of a Random Polling Dynamic Load Balancing protocol has been examined. Finally, the PN models are validated by a MPI benchmarking program running on the real multiprocessor system. The advantages of new pipelined communication designs as well as the benefits of PN approach for evaluating and developing high performance parallel mesh refinement algorithms are demonstrated.

This thesis analyzes procedure of generation of hyperbolic grids formulated by two constraints, which specify grid orthogonality and cell volume. The procedure was applied on a wide range of geometries and high quality two and three dimensional hyperbolic grids were generated by using grid control and smoothing procedures, which supply grid clustering in all directions and prevent grid deformation (grid shock), respectively.

Thesis (Ph. D.)--Aerospace Engineering, Georgia Institute of Technology, 2008. / Committee Chair: Tsiotras, Panagiotis; Committee Member: Calise, Anthony J.; Committee Member: Egerstedt, Magnus; Committee Member: Prasad, J. V. R.; Committee Member: Russell, Ryan P.; Committee Member: Zhou, Hao-Min.

Thesis (M.S.)--Rochester Institute of Technology, 1990. / Typescript. Includes bibliographical references.

Thesis (Ph. D.)--University of Washington, 1999. / Vita. Includes bibliographical references (p. 165-171).

Thesis (M.S.)--Brigham Young University. Dept. of Mathematics, 2004. / Includes bibliographical references (p. 74-75).

Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: genetic algorithms; FEM mesh elements; mapping. Includes bibliographical references (p. 52-53).

"[pi]" appears as a symbol in title. Includes bibliographical references (p. 35-41).