A novel approach for part family formation using K-means algorithm

Abstract

The reconfigurable manufacturing system (RMS) is the next step in manufacturing, allowing the production of any quantity of highly customized and complex parts together with the benefits of mass production. In RMSs, parts are grouped into families, each of which requires a specific system configuration. Initially system is configured to produce the first family of parts. Once it is finished, the system will be reconfigured in order to produce the second family, and so forth. The effectiveness of a RMS depends on the formation of the optimum set of part families addressing various reconfigurability issues. The aim of this work is to establish a methodology for grouping parts into families for effective working of RMS. The methodology carried out in two phases. In the first phase, the correlation matrix is used as similarity coefficient matrix. In the second phase, agglomerative hierarchical K-means algorithm is used for the parts family formation resulting in an optimum set of part families for reconfigurable manufacturing system.

Keywords

Reconfigurable manufacturing system (RMS) / Part family formation / K-means algorithm / Similarity coefficient / Correlation matrix

References

[1] Koren Y, Jovane F, Heisel U, et al. Reconfigurable manufacturing systems. Ann CIRP, 1999, 48(2): 527-540.

[2] Lokesh K, Jain PK. A model and optimization approach for reconfigurable manufacturing system configuration design. Int J Prod Res, 2011, 50(12): 3359-3381.

[3] Abdi MR, Labib AW. Grouping and selecting products: the design key of reconfigurable manufacturing systems (RMSs). Int J Prod Res, 2004, 42(3): 521-546.

[4] Galan R, Racero J, Eguia I, et al. A systematic approach for product families formation in reconfigurable manufacturing systems. Robot Comp Integr Manuf, 2007, 23(5): 489-502.

[5] Mahesh O, Srinivasan G. Incremental cell formation considering alternative machines. Int J Prod Res, 2002, 40(14): 3291-3310.

[6] McAuley J. Machine grouping for efficient production. Prod Eng, 1972, 51(2): 53-57.

[7] Mosier CT, Taube L. Weighted similarity measure heuristics for the group technology machine clustering problem. Int J Manag Sci, 1985, 13: 577-583.

[8] Wei JC, Kern GM. Commonality analysis: a linear clustering algorithm for group technology. Int J Prod Res, 1989, 12: 2053-2062.

[9] Sarker BR, Islam KM. Relative performances of similarity and dissimilarity measures. Comp Ind Eng, 1999, 37: 769-807.

[10] Vakharia AJ, Wemmerlöv U. A comparative investigation of hierarchical clustering techniques and dissimilarity measures applied to the cell formation problem. J Oper Manag, 1995, 13: 117-138.

[11] Zalik KR. An efficient K-means clustering algorithm. Pattern Recognit Lett, 2008, 29: 1385-1391.

[12] Jain AK. Data clustering: 50 years beyond K-means. Pattern Recognit Lett, 2010, 31: 651-666.

[13] Tsai CF, Hsu YF, Lin CY, et al. Intrusion detection by machine learning: a review. Expert Syst Appl, 2009, 36: 11994-12000.

[14] Al-Sultan KS. A hard clustering approach to the part family formation problem. Prod Plann Control, 1997, 8(3): 231-236.

[15] Shehroz SK, Ahmad A. Cluster centre initialization algorithm for K-means clustering. Pattern Recogni Letter, 2004, 25: 1293-1302.