Search for content and authors

Genetic algorithms for cluster optimization: general aspects

Roy L. Johnston 

School of Chemistry, University of Birmingham, Edgbaston, Birmingham B152TT, United Kingdom


In this lecture, I will start by giving a short overview of the field of cluster science and the reasons for using genetic algorithms for global optimization in this field.  I will then present a brief history of the application of genetic algorithms to various types of clusters.  Finally, I will describe the development of our own genetic algorithm (the Birmingham Cluster Genetic Algorithm) for cluster geometry optimization - giving examples of its application to a variety of different types of clusters.

During the lecture, I will emphasize how the nature of the problem or system being studied influences our choice of genetic algorithm methodology and operators and I will introduce a number of ways in which the basic genetic algorithm can be improved - for example by developing hybrid search methodologies to enable local as well as global searching.  I will also talk about potential pitfalls and possible future directions.


Legal notice
  • Legal notice:

Related papers

Presentation: Invited at E-MRS Fall Meeting 2007, Genetic algorithms for beginners, by Roy L. Johnston
See On-line Journal of E-MRS Fall Meeting 2007

Submitted: 2007-06-20 13:35
Revised:   2009-06-07 00:44