MURAL - Maynooth University Research Archive Library

    Improving Results of Differential Evolution Algorithm

    Sawant, Rushikesh (2015) Improving Results of Differential Evolution Algorithm. Masters thesis, National University of Ireland Maynooth.

    Download (1MB) | Preview

    Share your research

    Twitter Facebook LinkedIn GooglePlus Email more...

    Add this article to your Mendeley library


    Optimisation problems are of prime importance in scientific and engineering communities. Many day-to-day tasks in these fields can be classified as optimisation problems. Due to their enormous solution spaces, optimisation problems frequently lie in class NP. In such cases, engineers and researchers have to rely on algorithms and techniques that can find sub-optimal solutions to these problems. One of the most dependable algorithms for numerical optimisation problems is Differential Evolution (DE). Since its introduction in the mid 1990’s, DE has been on the fore front when it comes to applicability of optimisation algorithms to variety of real-parameter optimisation problems. This popularity of DE has driven intensive research to further improve its capability to find optimal solutions. In this thesis we present a variant of DE to produce improved solutions with greater reliability. In doing so, we introduce a novel strategy to incorporate ancestral vectors into the optimisation process. We show that a controlled introduction of ancestral vectors into the optimisation process has a generally positive influence on convergence rate of the algorithm. Evaluation of the proposed algorithm forms a major part of this work, as an empirical evidence serves to demonstrate the performance of stochastic algorithms. The resulting implementation of the algorithm is made available as an open source software along with its reference manual.

    Item Type: Thesis (Masters)
    Additional Information: Taught Masters Thesis for the Erasmus Mundus MSc in Dependable Software Systems
    Keywords: Improving Results; Differential Evolution Algorithm;
    Academic Unit: Faculty of Science and Engineering > Computer Science
    Item ID: 7091
    Depositing User: IR eTheses
    Date Deposited: 04 May 2016 11:07
      Use Licence: This item is available under a Creative Commons Attribution Non Commercial Share Alike Licence (CC BY-NC-SA). Details of this licence are available here

      Repository Staff Only(login required)

      View Item Item control page


      Downloads per month over past year

      Origin of downloads