Optimising the time-based Design Structure Matrix using a Divide and Hybridise Algorithm

Cook, I.; Coates, G.

Abstract

Product design and development processes consist of inter-related activities required to be undertaken in an appropriate sequence to reduce the need for iteration and increase opportunities for concurrency. The design structure matrix (DSM) is one of several modelling tools used to represent activities and the dependencies between them. Many algorithms have been developed and applied to the DSM with the aim of determining a near-optimal sequence of activities in terms of a range of objectives. In this paper, an enhanced genetic algorithm (GA), referred to as the divide and hybridise algorithm (DaHA), is applied to sequence the DSM with the objectives of minimising iteration and maximising concurrency simultaneously. The DaHA includes a new form of niching, which involves a population being divided into sub-populations, creating an opportunity for each to locate their own local optimum. Sub-populations are then hybridised to explore the solution space between these optima. Furthermore, a new ordinal-based selection method is presented, which encourages diversity and enables a more thorough exploration of the solution space than existing ordinal-based methods. Finally, the DaHA has been compared with several other algorithms from the literature and found to give superior, or at least equal, results when sequencing the DSM.

Citation

Cook, I., & Coates, G. (2016). Optimising the time-based Design Structure Matrix using a Divide and Hybridise Algorithm. Journal of Engineering Design, 27(4-6), 306-332. https://doi.org/10.1080/09544828.2016.1146236

• Journal Article Type: Article
• Acceptance Date: Jan 21, 2016
• Online Publication Date: Feb 22, 2016
• Publication Date: May 1, 2016
• Deposit Date: Jan 26, 2016
• Publicly Available Date: Feb 22, 2017
• Journal: Journal of Engineering Design
• Print ISSN: 0954-4828
• Electronic ISSN: 1466-1837
• Publisher: Taylor and Francis Group
• Peer Reviewed: Peer Reviewed
• Volume: 27
• Issue: 4-6
• Pages: 306-332
• DOI: https://doi.org/10.1080/09544828.2016.1146236
• Public URL: https://durham-repository.worktribe.com/output/1394201