JCell is a framework for working mainly with cellular genetic algorithms (cGAs), but also it has implemented steady-state GAs, generational GAs, and distributed GAs (only a sequential version with ssGAs in the islands). The design of JCell allows the user to implement any metaheuristic, so we would be glad if you provide us your new algorithms (as well as operators, representations, etc.) implemented in JCell to add them to this repository.
JCell is coded in Java and provides to the user some of the most recent proposed techniques in the literature of cGAs (e.g., self-adaptive and hierarchical populations, anysotropic selection, a novel multi-objective cGA, etc.). The use of JCell is very simple, since it is only required the manipulation of a simple configuration file (some example configuration files are provided in the cfg directory). JCell allows the user to work in combinatorial optimization, integer programming, continuous optimization, and multi-objective environments, all this with or without constraints.
We consider JCell a really interesting and useful tool for future research, since it allows the combination of several new promising techniques recently published in the literature. Additionally, its careful design following the software engineering recommendations provides an intuitive code, allowing the user to easily make modifications and/or add new features to the framework. JCell is implemented in Java, a very well-known programming language that allows to execute our code in most computer platforms without any modification. It is possible because Java is an interpreted language, and thus a Java program can run in any kind computer having the Java Virtual Machine --the Java interpreter-- installed. JCell is compatible from version JDK 1.5 ahead.
JCell can be used both as a library that can be imported by our Java program or it can be directly executed using the file JCell.class. Klick the links below, for more information on the descpription of JCell or the use of its configuration file.
Please, note that this is a beta version of the library, and not all its possibilities have been tested. If you find errors, please send the configuration file to the administrator of this site.
DOWNLOAD via web the source code for JCell
Second option, download the Source Code typing the following command line:
svn checkout svn://svn.gforge.uni.lu/svn/jcell
Besides, we recommend you to visit:
- JCell Project Site: The first site to go to get more info.
- The cEAs web site: It is a large web site dedicated to cellular evolutionary algorithms.
- The VRP Web: All you need about the VRP (Vehicle Routing Problem): problem description, problem variants, instances, algorithms used for solving it, best known solutions, links, bibliography...
- jMetal: A framework for multiobjective programming. It contains really efficient and implementations of the state-of-the-art algorithms in the multi-objective domain, including cMOGA, a multi-objective cGA.
Recommended Book (New Release):
Cellular Genetic Algorithms E. Alba, B. Dorronsoro Series: Operations Research / Computer Science Vol. 42 250 p. 2008 Springer-Verlag ISBN: 978-0-387-77609-5 Table of Contents | Buy it! |
The author of this content is Bernabé Dorronsoro.
Contact by e-mail: bernabe at lcc.uma.es