Effect of Link Length, Population Size, and Mutation Rate on the Convergence of an Order-Based Genetic Algorithm
Stelson, K. and Oba, F.
The convergence of an order-based genetic algorithm with random selection and mutation is analyzed in this paper. A measure of difference between two links called `link distance' is defined and used to study the convergence of the genetic algorithm. The changes in the expected average link distance of a mating pool due to genetic drift and mutation are first derived separately and then combined to describe their joint effect on the convergence of the genetic algorithm. The iterations required to reach convergence for various link lengths, population sizes, and mutation rates are obtained. The effects of these parameters are discussed.
H. Wong and M. Leu, "Effect of Link Length, Population Size, and Mutation Rate on the Convergence of an Order-Based Genetic Algorithm," Proceedings of the Japan/USA Symposium on Flexible Automation, American Society of Mechanical Engineers (ASME), Jan 1996.
1996 Japan-USA Symposium on Flexible Automation Part 2
Mechanical and Aerospace Engineering
Keywords and Phrases
Convergence of Numerical Methods; Iterative Methods; Neural Methods; Optimization
Article - Conference proceedings
© 1996 American Society of Mechanical Engineers (ASME), All rights reserved.