A body of literature addresses the optimization of
production-distribution systems without the formulation
of MIP models. As an illustration, we briefly
present two of these papers. Burns et al. (1985)
present an analytic approach to optimize freight distribution
by truck from a supplier to many customers,
considering transportation and inventory
costs. Their approach compares two distribution
strategies, namely, shipping separate loads to each
customer, and dispatching one truck to several customer.
The analytical formulas developed by the
authors may be used to approximate distribution
costs, requiting less data in comparison to traditional
network modeling approaches. Clearly, this approach
is not applicable in the strategic design of global
supply chains, nor does it include facility location
decisions and production considerations.
Cohen and Lee (1988) develop a set of stochastic
submodels including the optimization of material
control operations, a serial production process, finished
goods stockpile, and a distribution process.
These submodels are linked and related to each other
using a set of variables. Under some assumptions,
the authors are able to optimize each submodel
independently. However, the resulting optimization
problem over all the submodels is untractable, and
thus the authors apply a heuristic procedure to obtain
good operating policies. Additionally, the problems
of facility location, capacity planning, and selection
of technology are not considered in the submodels
since they are assumed to be fixed. Although the
applicability of this approach in the design of GLS is
restricted, it appears to be a part of the normative
model framework by Cohen and Kleindorfer (1993).
Of the MIP models reviewed above, the model by
Cole (1995) is the only one that explicitly includes
stochastic safety stock as a customer service factor;
it appears to be the most comprehensive model for a
single-country environment. The model by Arntzen
et al. (1995) appears as the most comprehensive one
for an international supply chain system. In Table 1
we show the main characteristics of some selected
MIP models.
A body of literature addresses the optimization ofproduction-distribution systems without the formulationof MIP models. As an illustration, we brieflypresent two of these papers. Burns et al. (1985)present an analytic approach to optimize freight distributionby truck from a supplier to many customers,considering transportation and inventorycosts. Their approach compares two distributionstrategies, namely, shipping separate loads to eachcustomer, and dispatching one truck to several customer.The analytical formulas developed by theauthors may be used to approximate distributioncosts, requiting less data in comparison to traditionalnetwork modeling approaches. Clearly, this approachis not applicable in the strategic design of globalsupply chains, nor does it include facility locationdecisions and production considerations.Cohen and Lee (1988) develop a set of stochasticsubmodels including the optimization of materialcontrol operations, a serial production process, finishedgoods stockpile, and a distribution process.These submodels are linked and related to each otherusing a set of variables. Under some assumptions,the authors are able to optimize each submodelindependently. However, the resulting optimizationproblem over all the submodels is untractable, andthus the authors apply a heuristic procedure to obtaingood operating policies. Additionally, the problemsof facility location, capacity planning, and selectionof technology are not considered in the submodels
since they are assumed to be fixed. Although the
applicability of this approach in the design of GLS is
restricted, it appears to be a part of the normative
model framework by Cohen and Kleindorfer (1993).
Of the MIP models reviewed above, the model by
Cole (1995) is the only one that explicitly includes
stochastic safety stock as a customer service factor;
it appears to be the most comprehensive model for a
single-country environment. The model by Arntzen
et al. (1995) appears as the most comprehensive one
for an international supply chain system. In Table 1
we show the main characteristics of some selected
MIP models.
การแปล กรุณารอสักครู่..
