14
C. J. Thomas Renald et al.
Grouping efficacy:
τ =
o − e
o + v
(6)
where
o Quantity of 1’s in the matrix
e Quantity of exceptional elements
v Quantity of voids.
τ = 1 implies 0 voids and 0 exceptional elements (perfect clustering)
τ = 0 implies all 1’s are outside the diagonal blocks.
w Weight factor
o Amount of 1’s in matrix
e Amount of exceptional elements
v Quantity of voids
M Quantity of machines
P Quantity of components
η = 1 implies 0 voids and 0 outstanding elements (perfect clustering)
e 0 number of exceptional elements.
2 Results and Discussions
From the solutions obtained, it is conformed that altered-one link clustering method
gives the better performance for the above problem. The result has two exceptional
elements, three voids, grouping efficiency is 91.33% and grouping efficacy is 82.14%.
Weighing factor was selected as 0.412, given by Mukhopadhyay et al. [36]. Thus, it
was identified that MOD-SLC formulae gives better result than the other two in this
case study. The results obtained from the three-cell formation algorithms are drawn
graphically from Figs. 2, 3, 4, 5 and 6. Table 1 shows the computational results
(Figs. 7, 8 and 9).
3 Graphical Results of Algorithm Versus Evaluating
Parameters
1. Fig. 2 shows the improved performance of MOD-SLC algorithm than ROC-2
and DCA and gives smaller number of exceptional elements of 2.
2. From Fig. 3, it is clear that MOD-SLC provides minimum number of voids than
other two algorithms.