A generator for multimodal test functions with multiple global optima

Abstract

The topic of multimodal function optimization, where the aim is to locate more than one solution, has attracted a growing interest especially in the evolutionary computing research community. To experimentally evaluate the strengths and weaknesses of multimodal optimization algorithms, it is important to use test functions representing different characteristics and of various levels of difficulty. However, the available selection of multimodal test problems with multiple global optima is rather limited at the moment and no general framework exists. This paper describes our attempt in constructing a test function generator to allow the generation of easily tunable test functions. The aim is to provide a general and easily expandable environment for testing different methods of multimodal optimization. Several function families with different characteristics are included. The generator implements new parameterizable function families for generating desired landscapes and a selection of well known test functions from literature, which can be rotated and stretched. The module can be easily imported to any optimization algorithm implementation compatible with C programming language. © 2008 Springer Berlin Heidelberg.