Evaluating the performance of the Similarity Coefficient Measures in the Feasibility Assessment of Cellular Manufacturing

Abstract

Cellular Manufacturing (CM) system is a modern and effective manufacturing system that considered as a best application of Group Technology (GT) principles. To apply CM in the real life, three issues should be counted (i) Feasibility Assessment (FA) (ii) design of CM (iii) application processes. FA is an important issue to evaluate the data of the job shop system and predict the number of machine cells before the designing and applying of CM. In the previous literature, there are a large number of Similarity Coefficients (SCs) have been proposed to solve the Cell Formation Problem (CFP). These SCs have different accuracy and the comparative studies that done to identify the performance of these SCs are very limited and focusing only on the design issue of CM. Also in the past literature, a comparative study on evaluating the performance of different SCs in the FA is overlooked. The novelty of the present study is to identify the performance of different SCs in the FA of CM. In this paper, 20 General Purposes Similarity Coefficients (GPSCs) and problem oriented SCs have been evaluated in the FA. A set of thirty test problems selected from literature are examined in this study. The differences in the number of machine cells that produced by using each SC comparing with the majority of the SCs are identified and used as evaluation criteria. The results show that the GPSCs and problem oriented SCs can be classified into two classes: the first class considered as an efficient GPSCs which include: Sorenson, Rusell and Rao, Ochiai, Dot-product, However the second class is inefficient which consists four GPSCs: Sokal and Sneath 4, Yule, Hamann and Phi.
Copyright © 2014 Praise Worthy Prize - All rights reserved.