An OOP MATLAB Extensible Framework for the Implementation of Genetic Algorithms. Part I: The Framework

Genetic algorithms are already a well-established method for structural or general optimization. There is a large number of libraries and frameworks available assisting the researchers and engineers alike to implement genetic algorithm codes. However, there isn't a satisfactory tool available for the de facto engineering programming language: MATLAB. The existing tools only use procedural programming and offer some flexibility but not to the extent an object oriented approach can provide. This paper presents the work of the authors regarding the design and implementation of a tool aimed to fill this void. The result is a fully object oriented framework for MATLAB, similar to a toolbox. This offers the general layout of any genetic algorithm, along with typical implementations. Its architecture is designed to offer full flexibility and extensibility. The base layout allows researchers to easily study, test and implement new or improved genetic operators or concepts, while the implementation of basic algorithms offers the practitioners a ready to use tool. The first part of the paper presents in some detail the concepts and structure of the framework, while the second part deals with a case study to illustrate the effectiveness of the approach. (C) 2015 The Authors. Published by Elsevier Ltd.