Graphics Programs Reference
In-Depth Information
M
1
S1
S2
M
2
LU
S3
S5
S4
M
3
Figure 8.8: Schematization of a FMS with continuous transportation system
A system using continuous transportation — In this example we
assume that type a parts must be processed by machine M
1
and machine
M
2
in sequence, while type b parts must be processed by machine M
2
and
machine M
3
in sequence. We also assume that no I/O buffer exists; the
parts waiting for a machine to become available queue up on the transport
path. The continuous transport system allows the flow of parts from LU to
the three machines in sequence and to LU again. When a part reaches the
segment of transport positioned in front of a machine, it either waits in that
position, if the service of that machine is required, or moves ahead. To model
the transport system with a GSPN, it must first be discretized, i.e., divided
into segments each capable of holding a pallet loaded with a part. A part
can move from a segment to the following segment only if the latter is free.
Each segment can thus be modelled by a place representing its availability,
and by two places representing the presence of either a type a or a type b
part in it. In Fig.
8.8
the continuous transport system is depicted: it has
been divided into five segments, S1,...,S5. Machines M
1
, M
2
and M
3
can
be accessed from segments S1, S3 and S5 respectively. Fig.
8.9
illustrates
the GSPN submodel of the transport system: the submodels representing a
segment in front of a machine contain one exiting arc towards the machine
and one entering arc from the machine.
Fig.
8.10.
An observation on this net is necessary: when a part leaves the
system, its identity is forgotten and it is replaced by a new part of unknown
type; the new part enters the system, and after its loading on a pallet,
it probabilistically receives a type identification by firing either transition
part
a
or transition part
b
. This is a “compact” representation of the input
Search WWH ::
Custom Search