Platform design variable identification for a product family using multi-objective particle swarm optimization

Seung Ki Moon, Kyoung Jong Park, Timothy W. Simpson

Research output: Contribution to journalArticlepeer-review

59 Scopus citations


The variability of products affects customers' satisfaction by increasing flexibility in decision-making for choosing a product based on their preferences in competitive market environments. In product family design, decision-making for determining a platform design strategy or the degree of commonality in a platform can be considered as a multidisciplinary optimization problem with respect to design variables, production cost, company's revenue, and customers' satisfaction. In this paper, we investigate evolutionary algorithms and module-based design approaches to identify an optimal platform strategy in a product family. The objective of this paper is to apply a multi-objective particle swarm optimization (MOPSO) approach to determine design variables for the best platform design strategy based on commonality and design variation within the product family. We describe modifications to apply the proposed MOPSO to the multi-objective problem of product family design and allow designers to evaluate varying levels of platform strategies. To demonstrate the effectiveness of the proposed approach, we use a case study involving a family of General Aviation Aircraft. We show that the proposed optimization algorithm can provide a proper solution in product family design process through experiments. The limitations of the approach and future work are also discussed.

Original languageEnglish (US)
Pages (from-to)95-108
Number of pages14
JournalResearch in Engineering Design
Issue number2
StatePublished - Apr 2014

All Science Journal Classification (ASJC) codes

  • Civil and Structural Engineering
  • Architecture
  • Mechanical Engineering
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Platform design variable identification for a product family using multi-objective particle swarm optimization'. Together they form a unique fingerprint.

Cite this