Hierarchical system concept generation

Rosenstein David, Yoram Reich

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The most important stage in a product life cycle is the conceptual design which involves uncertainty but also opportunity. The SOS (Subjective Objective System) method of generating product design alternatives [1] is expanded here to introduce more information to help reduce uncertainty and explore design solutions to better suite customer requirements, market conditions, and the use of current technology. Hierarchical SOS uses two levels of building blocks: the 1 st level contains major building blocks (the same as in regular SOS) and the 2 nd level contains slightly more detailed articulations of the building blocks of the 1 st level. The product requirements are cascaded to the 2 nd level and are translated to secondary targets. The search of the best product design alternatives is done by a genetic algorithm (GA). As expected, the new information introduced and manipulated at the 2 nd level turns out to be critical for creating good conceptual designs of systems.

Original languageEnglish
Title of host publicationICED 11 - 18th International Conference on Engineering Design - Impacting Society Through Engineering Design
Pages24-34
Number of pages11
EditionPART 2
StatePublished - 2011
Event18th International Conference on Engineering Design, ICED 11 - Copenhagen, Denmark
Duration: 15 Aug 201118 Aug 2011

Publication series

NameICED 11 - 18th International Conference on Engineering Design - Impacting Society Through Engineering Design
NumberPART 2
Volume10

Conference

Conference18th International Conference on Engineering Design, ICED 11
Country/TerritoryDenmark
CityCopenhagen
Period15/08/1118/08/11

Keywords

  • Conceptual design
  • Genetic algorithm
  • NSGA-II
  • Subjective objective design
  • Target cascading

Fingerprint

Dive into the research topics of 'Hierarchical system concept generation'. Together they form a unique fingerprint.

Cite this