期刊名称：Journal of Information Science and Technology
印刷版ISSN：1545-0287
出版年度：2010
卷号：7
期号：2
页码：36-57
语种：English
出版社：International Medical Journal Management and Indexing System
摘要：Evolutionary Strategies (ES) are effective forms of Evolutionary Algorithmsthat enable solving optimization problems. Effective use of ES algorithmhas been made in numerous fields. Major optimization problems oftoday possess a very complex fitness landscape with numerous modalities.The optimization in these complex landscapes is much more difficult as it ispossible only to explore a relatively small section of the entire landscape.Also the fitness function behaves in a very sensitive manner with a largeamount of change for small changes in the parameter values. We hencepropose a hierarchical ES to optimally explore the fitness landscape and returnthe optima. The inner or the slave ES is controlled by a controlling algorithmor the master. The master has a number of slave ES, each trying to finda solution at some different part of the complex high dimensional fitness landscape.Each ES tries to find the optimal point in its local surroundings. Hencethe variable step size is initially kept low. As the iterations of the master increase,we keep reducing the number of ESs and increase the step size togive it a global nature. This is the local to global nature search performed bythe algorithm. Since the fitness landscape is complex, the master mutatesthe locations of the ESs and adds new ESs (deleting the non-optimal ones)as iterations or generations proceed. The novelty of the suggested approach lies in the tradeoff between the search for global optima and convergence tolocal optima that can be controlled between the two hierarchies. Experimentalanalysis shows that the proposed algorithm gives a decent performance insimple optimization problems, but a better performance as we increase thecomplexity, when compared with the conventional Genetic Algorithm, ParticleSwarm Optimization and conventional ES.
