Metaheuristic Design Patterns: New Perspectives for Larger-Scale Search Architectures

Krzysztof Krawiec, Christopher Simons, Jerry Swan, John R. Woodward

Research output: Chapter in Book/Report/Conference proceedingOther chapter contribution


Design patterns capture the essentials of recurring best practice in an abstract form. Their merits are well established in domains as diverse as architecture and software development. They offer significant benefits, not least a common conceptual vocabulary for designers, enabling greater communication of high-level concerns and increased software reuse. Inspired by the success of software design patterns, this chapter seeks to promote the merits of a pattern-based method to the development of metaheuristic search software components. To achieve this, a catalog of patterns is presented, organized into the families of structural, behavioral, methodological and component-based patterns. As an alternative to the increasing specialization associated with individual metaheuristic search components, the authors encourage computer scientists to embrace the ‘cross cutting' benefits of a pattern-based perspective to optimization algorithms. Some ways in which the patterns might form the basis of further larger-scale metaheuristic component design automation are also discussed.
Original languageEnglish
Title of host publicationHandbook of Research on Emergent Applications of Optimization Algorithms
EditorsPandian Vasant, Sirma Zeynep Alparslan-Gok, Gerhard-Wilhelm Weber
Place of PublicationHershey, PA, USA
Publisher IGI Global
Number of pages36
Publication statusPublished - 2018

Cite this