Anda di halaman 1dari 3

vevenkatakrisVenkatakrishnan Siruvalore Lakshminaraynan, Masters I.

E
1
Flexible Manufacturing System
Assignment-2
Cellular manufacturing is a part of group technology, in which similar parts are
grouped together into families and machines are grouped together so that more
than one part family can be machined by a single machine cell. In cellular
manufacturing, parts are not produced continuously but intermittently produced in
small lot sizes. CM does not face the problem like frequent setups, excessive
in-process inventories, long through-put times, complex planning and control
functions, and provides the basis for implementation of manufacturing techniques
such as just-in-time (JIT) and flexible manufacturing systems (FMS).

The machine cell formation is performed by three methods: 1. Mathematical
Programming (MP) models, 2. (Meta-)Heuristic Algorithms (HA) 3. Similarity
Coefficient Methods (SCM). The Similarity coefficient method (SCM) is used for
comparing the similarity and diversity of sample sets. The SCM is the application
of cluster analysis to cell formation. The advantages of SCM over Mathematical
programming and Meta-Heuristic Algorithms are 1) development of new model
from existing model is easy, 2) common way exists for solving a complex system
into many smaller less complex system.

Taxonomy in SCM exists which is the basis for grouping and diversification. The
5-level hierarchy starting from 0-4 where, 0 is the root of the taxonomy. The level
zero denotes the similar coefficients. Level 1-problem oriented or general. Level-2,
Binary data based or production information based. The production information is
further classified into 1.alternate process, 2. Operation sequence, 3. Weight factors
and 4. Others. The Weight factors are classified in the level-5 into Production
volume and operation time.

Production information based similarity coefficient includes alternate processing
routes, operation sequence and weight factors. The alternate processing routes was
proposed and used by Gupta (1993) which is further developed by Won and Kim
(1997). The basic idea is that in the final solution only one process routing is
selected for each part.
The operation sequence is the order in which parts enter and leave the machine.
Selvam and Balasubramanian (1985) proposed that the value of similarity
coefficient is obtained directly by the production volume of parts moving between
machines. Seifoddini (1987/1988) modified Jaccard similarity coefficient to take
into account the production volume of parts moving between machine pairs. Gupta
and Seifoddini (1990) proposed a similarity coefficient incorporating operation
vevenkatakrisVenkatakrishnan Siruvalore Lakshminaraynan, Masters I.E
2
sequence, production volume and operation time simultaneously. From the
definition, each part that is processed by at least one machine from a pair of
machines leads to a similarity coefficient value. A part that is processed by both
machines increases the coefficient whereas, a part that is processed on one machine
tends to reduce it. Balasubramanian and Panneerselvam (1993) developed a
similarity coefficient which needs the input data of: (1) operation sequences of
parts; (2) additional cell arrangements; (3) production volume per day and the bulk
factor; (4) guidelines for computing excess moves; (5) actual cost per move. The
similarity coefficient obtained is applied to a p-median to group parts into part
families.

Weight factors is a binary data based similarity coefficient. The weight factors
considered are production volume and operation time.

The tested data sets are classified into two distinct types: from literature and
generated deliberately Let us consider the deliberately generated data types: 1.
Ratio of non-zero elements in cell (REC), 2. Ratio of exceptions. REC has three
densities that are called Problem Density (PD), non-zero elements inside cells
Density (ID) and non-zero elements outside cells Density (OD). The densities are
used to find the REC. If REC is bigger than 1, current machine capacity cannot
response to the productive requirements of parts. Thus, additional machines need
to be considered. Therefore, REC can be used as a means to assess the capacity of
machines. RE is used to judge the goodness of machine-part cells and
distinguish well-structured problems from ill-structured problems.

There are many clustering procedures of which most widely used procedures
include single linkage clustering (SLC) algorithm, complete linkage clustering
(CLC) algorithm and average linkage clustering (ALC) algorithm. SLC has the
advantage of Simplicity; Minimal computational requirement; Tends to minimize
the degree of adjusted machine duplication. It also has the tendency to chain;
Leads to the lowest densities and the highest degree of single part cells which
makes it a drawback. CLC has Simplicity; Minimal computational requirement as
an advantage and also it is considered as a worst procedure (Vakharia &
Wemmerlv, 1995; Yasuda & Yin, 2001). ALC is performed as the best procedure;
Produces the lowest degree of chaining; Leads to the highest cell densities;
Indifferent to choice of similarity coefficients; few single part cells however it
requires high level of machine duplication. The performance measures of each
SCM is determined based on Number of exceptional elements (EE), Grouping
efficiency, Group efficacy, Machine utilization index (Grouping measure, GM),
Clustering measure (CM), Grouping index (GI), Bond energy measure (BEM),
vevenkatakrisVenkatakrishnan Siruvalore Lakshminaraynan, Masters I.E
3
Grouping capability index (GCI), Alternative routeing grouping efficiency (ARG
efficiency),

Anda mungkin juga menyukai