Document Type thesis Author Name Dunkelberg, Jr., John S. URN etd-0104101-102839 Title FEM Mesh Mapping to a SIMD Machine Using Genetic Algorithms Degree MS Department Computer Science Advisors Dr. Lee A. Becker, Advisor Dr. William R. Michalson, Co-Advisor Dr. Robert E. Kinicki, Department Head Keywords genetic algorithm FEM mesh elements mapping Date of Presentation/Defense 1996-05-01 Availability unrestricted
The Finite Element Method is a computationally expensive method used to perform engineering analyses. By performing such computations on a parallel machine using a SIMD paradigm, these analyses’ run time can be drastically reduced. However, the mapping of the FEM mesh elements to the SIMD machine processing elements is an NP-complete problem. This thesis examines the use of Genetic Algorithms as a search technique to find quality solutions to the mapping problem. A hill climbing algorithm is compared to a traditional genetic algorithm, as well as a "messy" genetic algorithm. The results and comparative advantages of these approaches are discussed.
Files appendix-a.zip appendix-b.zip appendix-c.zip appendix-d.zip dunkelberg.pdf
Browse by Author | Browse by Department | Search all available ETDs
Questions? Email firstname.lastname@example.org