Date of Award

Spring 3-9-2012

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computer Science

First Advisor

Yan-Qing Zhang

Second Advisor

Raj Sunderraman

Third Advisor

YingShu Li

Fourth Advisor

Yichuan Zhao

Abstract

The Genetic Algorithm (GA) is a popular global search algorithm. Although it has been used successfully in many fields, there are still performance challenges that prevent GA’s further success. The performance challenges include: difficult to reach optimal solutions for complex problems and take a very long time to solve difficult problems. This dissertation is to research new ways to improve GA’s performance on solution quality and convergence speed. The main focus is to present the concept of shadow price and propose a two-measurement GA. The new algorithm uses the fitness value to measure solutions and shadow price to evaluate components. New shadow price Guided operators are used to achieve good measurable evolutions. Simulation results have shown that the new shadow price Guided genetic algorithm (SGA) is effective in terms of performance and efficient in terms of speed.

DOI

https://doi.org/10.57709/2721649

Share

COinS