Thesis (Ph.D.)--University of Delaware, 2009. / Principal faculty advisor: Joseph I. Daniel, Dept. of Economics. Includes bibliographical references.

Thesis (Ph. D.)--York University, 2001. / Typescript. Includes bibliographical references (leaves 107-115). Also available on the Internet. MODE OF ACCESS via web browser by entering the following URL: http://wwwlib.umi.com/cr/yorku/fullcit?pNQ67907.

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

Thesis (M.S.)--West Virginia University, 2004. / Title from document title page. Document formatted into pages; contains xii, 112 p. : ill. Includes abstract. Includes bibliographical references (p. 95-99).

Thesis (M. Phil.)--University of Hong Kong, 1993.

Thesis (M.S.)--Rochester Institute of Technology, 2005. / Typescript. Includes bibliographical references (leaves 81-83).

Simulated Annealing is a general purpose combinatorial optimization technique which has been applied to many problems in VLSI design. In essence, simulated annealing is Monte Carlo iterative improvement with the ability to conditionally accept uphill moves. The notion of a cooling schedule is common to all simulated annealing implementations. A cooling schedule can be thought of as simulated annealing's control mechanisms. Experiential work has been done on estimating the cost of an optimal solution to some combinatorial optimization problem instances. Such an estimate can be used to determine termination criteria for general purpose optimization techniques such as iterative improvement or simulated annealing. We have extended this idea and designed a complete simulated annealing general cooling schedule based on the cost of an optimal solution to a problem instance. We call the resultant schedule an extended goal-directed general cooling schedule. One of the major problems with simulated annealing is its long computation times. This problem can be addressed by first using a fast heuristic to find a good initial configuration and then applying simulated annealing. This approach is called Simulated Sintering. To exploit the potential of simulated sintering one needs an appropriate general cooling schedule. The extended goal-directed cooling schedule is equally applicable to simulated annealing and simulated sintering. To date, no one cooling schedule has proven suitable for all optimization problem instances. In our view, no such cooling schedule exits. Consequently, we have attempted to identify the type of problem best suited to optimization by simulated annealing and simulated sintering using the extended goal-directed schedule. We have applied the extended goal-directed schedule to standard-cell placement and floorplanning problems using both simulated annealing and simulated sintering. Within this context, we have compared the performance of the extended goal-directed schedule to other published schedules. Our results indicate that in terms of layout quality, the extended goal-directed schedule performs as well or better than the other schedules. In this dissertation, we have developed a new general cooling schedule. Our evaluation of the extended goal-directed schedule suggests that it is a useful research contribution in the area of simulated annealing algorithms. / Graduate

This thesis investigates the effect of neighborhood structure on simulated annealing, a random search algorithm that has the ability to search through a sequence of local optimal solutions and return the globally optimal solution. Neighborhood sizes of one, two, three, four and N (equal to the size of the problem) decision variable alterations have been experimentally tested on the following problem types: Quadratic Assignment problem; Quadratic Selection problem; Stochastic Optimization problem; and Traveling Salesman problem. Empirical results indicate that a smaller neighborhood size is better. However, for some instances, a neighborhood size of one larger than the smallest defined neighborhood size performed best.

The data used for the construction of genome maps is imperfect, therefore the mapping of a physically linear structure must take place in a very uneven feature space. As the number of genes to be ordered grows, it appears to be impractical to use exhaustive search techniques to find the optimal mapping. In this paper we compare genetic algorithms and simulated annealing, two methods that are widely believed to be well-suited to non-smooth feature spaces, and find that the genetic algorithm approach yields superior results. Here we present performance profiles of comparable implementations of both genetic algorithms and simulated annealing. We have translated the problem to a form comparable to the shortest-path problem and found that the ability of a genetic algorithm to combine different partial solutions seems to be responsible for its superiority over the simulated annealing method. This is because in the genome mapping problem, as in the Traveling Salesman Problem, good solutions tend to be rather sparse and because optimal subtours tend to be components of nearly optimal tours. / Graduation date: 1994

Thesis (M.S.)--West Virginia University, 2001. / Title from document title page. Document formatted into pages; contains x, 133 p. : ill. Includes abstract. Includes bibliographical references (p. 88-94).