Designing loudspeaker by ensemble of composite differential evolution ingredients

Xin Zhang, Xiu Zhang, Siu Lau Ho, Weinong Fu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

4 Citations (Scopus)


Design of electromagnetic devices has multimodal, multidimensional, and constrained characteristics. Metaheuristic approaches are good choices for tackling these design problems owing to their simulation-based property. As many electromagnetic design problems require a long computing time to solve (even on modern computers) and many heuristic approaches have been created, the major goal of this paper is to improve the effectiveness and robustness of existing approaches. This paper proposes an algorithm with the ensemble of two composite differential evolution (DE) ingredients. One ingredient is biased toward exploration and the other is biased toward exploitation. The probability of choosing which ingredient to search for new solutions is adaptively updated based on the previous performance of each ingredient. The algorithm is applied to solve a loudspeaker design problem with promising performance when compared with DE, artificial bee colony, two improved artificial bee colony algorithms, and a randomly choosing ingredients algorithm.
Original languageEnglish
Article number6971651
JournalIEEE Transactions on Magnetics
Issue number11
Publication statusPublished - 1 Nov 2014


  • Design optimization
  • differential evolution (DE)
  • ensemble method
  • loudspeaker

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Electrical and Electronic Engineering

Cite this