Geoscience Reference
In-Depth Information
sometimes referred to as
site-generation models
since the generation of appropriate
sites is left to the model in hand. On the other hand, in a discrete space facilities may
only be located at some predefined points. For this reason discrete space models can
also be referred to as
site-selection models
since the choice is limited within a set
of known candidates. Using network based models the choice may be restricted to
nodes or to any point of the network (node and/or arc). When a simultaneous choice
of nodes and arcs is required, the problem is usually referred to as a
network design
problem
. An example of this class is the so-called
corridor location problems
where
routes of arcs connecting two points have to be located. The characteristics of the
location space and the specific application generally drive the adoption of a metric
that is used to measure distances between elements of the space (facilities and/or
demand points).
Facilities
The term
facility
is used to denote an object to be located in order
to optimize the interaction with other pre-existing objects. Classical examples of
facilities are industrial or commercial structures (e.g., retail outlets, plants, ware-
houses, bank branches), public services sites (e.g., schools, hospitals, fire stations,
waste disposal sites), transportation and logistics infrastructures (e.g., terminals,
cross-dockings, metro stations, parking lots). Facilities are usually characterized by
attributes such as the number and the type of services they provide, their capacity,
their attractiveness, the costs associated with their establishment and operation.
Depending on the “intensity” of these attributes, facilities may produce certain
“effects” on a set of actors. If these effects are judged as positive, then facilities
are defined as “desirable”. For instance, this is the case of schools, public service
sites or metro stations where users generally wish to be as close to them as possible.
Otherwise they are considered “undesirable” as in the case of nuclear or chemical
plants, waste disposal sites or incinerators, airport or military installations and so on.
There also exist situations where facilities are partly desirable, partly undesirable
(e.g., commercial stores) as they produce some positive effects (i.e. accessibility to
services) as well as some negative ones (i.e. traffic congestion) on the surrounding
area.
A fundamental characteristic of a FLP is the number of new facilities to be
located. The simplest case is the
single-facility problem
when the position of only
one facility has to be determined, while the more general one is the
multi-facility
problem
in which the aim is to simultaneously locate more than one facility. The
number of facilities can be either pre-specified or a decision variable of the problem.
In the latter case, there may be restrictions on the minimum or the maximum number
of facilities to be located. The decision problem can also consider the possibility to
shut down existing facilities or to reposition some of the existing ones.
Demand
It represents the actors involved in the FLP. Depending on the kind of
service provided, they can be defined as customers, users, residents, population
centers and so on. Demand can be represented in continuous or in discrete fashion.
In the first case the demand area may be partitioned into sub-areas such that
within each sub-area it may be assumed that the demand is uniformly distributed.
