scispace - formally typeset
Search or ask a question
Author

Du Xin

Bio: Du Xin is an academic researcher. The author has contributed to research in topics: Realization (systems). The author has an hindex of 1, co-authored 1 publications receiving 1 citations.

Papers
More filters
Journal Article
TL;DR: Generally speaking, it takes lots of time to count the fitness in GA, and it continually generates a new generation that has lots of individuals so it is very prominent how to improve GA ‘s running speed.
Abstract: .With the application of the genetic algorithms(GA) deeply developed,the research of PGA(parallel genetic algorithms) and its realization become very important.Generally speaking,it takes lots of time to count the fitness in GA,and it continually generates a new generation that has lots of individuals.So it is very prominent how to improve GA ‘s running speed.Because of it’s inner parallel mechanism,it’s parallel process becomes very naturally resolvable method.

1 citations


Cited by
More filters
Proceedings ArticleDOI
18 Nov 2008
TL;DR: Four kinds of parallel models of parallel genetic algorithms, such as master- slave model, coarse- grained model, fine-grained model and mixed model, are generalized and evaluated and it is shown that parallel Genetic algorithms should go on further study in the future.
Abstract: With the application of the genetic algorithm (GA) deeply developed, the research of parallel genetic algorithm (PGA) and its realization become very important. Because of PGA inner parallel mechanism, its parallel process becomes a very naturally resolvable method. In this paper, four kinds of parallel models of parallel genetic algorithms, such as master- slave model, coarse-grained model, fine-grained model and mixed model, are simply generalized and evaluated. For every model, its characteristics are displayed. As for the existing problem to each model, the concerning parameters are illustrated in order to improve them. Then some main evaluation models of parallel genetic algorithms are presented. At the end, it is shown that parallel genetic algorithms should go on further study in the future.

18 citations