 |
    |
An Effective Real-Parameter Genetic Algorithm with Parent Centric Normal Crossover for Multimodal Optimisation
Pedro J. Ballester and Jonathan N. Carter
Imperial College London, Department of Earth Science and Engineering, RSM Building, Exhibition Road, London SW7 2AZ, UKp.ballester@imperial.ac.uk
j.n.carter@imperial.ac.uk
Abstract. Evolutionary Algorithms (EAs) are a useful tool to tackle real-world optimisation problems. Two important features that make these problems hard are multimodality and high dimensionality of the search landscape.
In this paper, we present a real-parameter Genetic Algorithm (GA) which is effective in optimising high dimensional, multimodal functions. We compare our algorithm with two previously published GAs which the authors claim gives good results for high dimensional, multimodal functions. For problems with only few local optima, our algorithm does not perform as well as one of the other algorithm. However, for problems with very many local optima, our algorithm performed significantly better. A wider comparison is made with previously published algorithms showing that our algorithm has the best performance for the hardest function tested.
LNCS 3102, p. 901 ff.
lncs@springer.de
© Springer-Verlag Berlin Heidelberg 2004