Impacts of Product Structural Characteristics on Modular Performance with Virtual DSM Data Mining

Shuai Wang,Deng-Zhuo Liu,Chao He,Guang-Yu Zou,Zhong-Kai Li
DOI: https://doi.org/10.1080/09544828.2023.2232983
2023-01-01
Journal of Engineering Design
Abstract:Increasingly fierce market competition and rapidly changing customer demand has led many manufacturing enterprises to actively adopt modular architecture for product design and development. This helps them to effectively control their production costs while maintaining the diversity of product variants. However, there are considerable differences between the structure of different products; the degree to which each product is suitable for adopting a modular architecture also varies. Here, we explored the impact of product structural features (the number of parts and the average number of connected nodes) on its modular performance, based on virtual product design structure matrix (DSM) data. We first propose a method for constructing a DSM of a virtual product and randomly generate a database of DSMs with different structural parameters. We then carry out hierarchical clustering and modularity evaluation on a DSM model to obtain the optimal modularity scheme and corresponding Q value. Finally, based on numerical simulation data, the relationship between the structural features of virtual products and their modular performance was analyzed. In addition, eight actual products were selected to verify the simulation results and ensure the credibility of these results.
What problem does this paper attempt to address?