Modular division method of complex product function-and-structure unit based on the improved fuzzy C-means clustering algorithm

Ma Shumei,Huang Lihui,Li Aiping,Duan Jianguo
DOI: https://doi.org/10.3969/j.issn.1671-3133.2012.02.003
2012-01-01
Abstract:To solve the problem in modular division of complex product function-and-structure,an improved fuzzy C-means(D-FCM)clustering algorithm for product function-and-structure was proposed.In this algorithm,distance matrix and F statistic were used to confirm the best cluster date,on this basis,fuzzy C-means clustering algorithm was used to get the result of clustering.Finally,a typical machine tool instance was given with the method of the modified fuzzy C-means clustering algorithm to test the method.At the same time,a system for machine tool granularity division was developed based on this method,which shows the validity and promise of the method mentioned above.Therefore it provides another numerical analysis method of machine tool for configuration design.
What problem does this paper attempt to address?