Discrete OptimizationOn solving large instances of the capacitated facility location problem
Introduction
In the capacitated facility location problem (CFLP), we have a set of customers with known demands of a certain commodity as well as a set of potential sites where facilities that satisfy the demands of the customers can be located. The capacity of the facility at each potential site is known. The problem is to determine the set of facilities to open as well as the manner in which the demand of each customer is satisfied without violating the capacity restriction of any facility, so that the total cost is minimized.
The polyhedral structures of the CFLP (and/or variants thereof) have been studied by Aardal et al., 1995, Aardal et al., 1996, Chen and Guignard, 1998, Deng and Simchi-Levi, 1993, Leung and Magnanti, 1989. Aardal (1998) provided separation algorithms and tested a strong fractional cutting-plane algorithm for the problem.
The present study has been motivated by a real-world exercise on capacitated facility location that the author has conducted; Sankaran and Raghavan (1997) reported an earlier stage of the study, in which the sites and customers respectively numbered 18 and 449.
Customer aggregation is a common device for managing the size of IP formulations for solving the CFLP and its variants (see for instance, Simchi-Levi et al., 2003); nevertheless, it necessitates an approximation of the true cost of servicing customer demand. For two continuous location models, the n-center and the n-median location problems introduced by Hakimi, 1964, Hakimi, 1965, Francis et al., 1996, Rayco et al., 1997 respectively developed aggregation schemes that exploited the underlying structures of the models. The CFLP is, in contrast, a discrete location model.
The disaggregate IP formulation of the CFLP, which contains variable-upper-bounding (VUB) constraints, is much tighter than the aggregate formulation. However, the VUB constraints are numerous, the more so for progressively large instances of the CFLP. In terms of applying a polyhedral approach for solving very large, real-world instances of the CFLP, a small subset (or even none) of the numerous VUB constraints could be included initially. The rest of the VUB constraints could be included subsequently on-the-fly, as and when needed, possibly along with any other strong fractional cutting-planes which might be yielded through separation algorithms, such as those featured in MINTO (Savelsbergh et al., 1994) and in Aardal (1998).
Two questions emerge from the above discussion. Given a large instance of the CFLP,
- 1.
How should one aggregate customers to compress the instance to a pre-specified size while controlling the extent to which the cost function is approximated?
- 2.
Which manageably small subset of the VUB constraints should one include at the outset of the solution of the instance of the CFLP so that the resulting duality gap is as small as possible?
Note the two questions speak to divergent means of pursuing the common goal of tractability. The first issue addresses the minimization of the error that customer aggregation introduces, while the second issue addresses the suspension of a large subset of VUB constraints that are not very critical to the closure of the duality gap.
In this note, we establish theoretical results that speak to the two issues listed above (proofs of all the results are presented in Appendix A). We begin by presenting the notation and the formulation. Subsequently, we present theoretical results, as also a heuristic for customer aggregation. Both sets of results, namely those pertaining to customer aggregation and those pertaining to the suspension of VUB constraints, are relevant in any real-world instance of facility location problems in which cities and towns define ‘customers’ and their ‘demands’. We present computational results concerning the first set of results with reference to a real-world variant of the CFLP.
Section snippets
The model
We use the following notation to describe the CFLP.
- m
number of sites
- n
number of customers
- D(j)
demand of customer j, j = 1, … , n
- c(i, j)
cost of serving one unit of demand at customer j from site i; in other words, the unit variable cost of serving customer j from site i (i = 1, … , m, j = 1, … , n)
- F(i)
fixed cost of locating a plant at site i, i = 1, … , m
- C(i)
capacity of a plant at site i, if built (i = 1, … , m)
- x(i, j)
proportion of the demand of customer j that is satisfied from site i (i = 1, … , m, j = 1, … , n)
- y(i)
a 0–1 decision
Results on customer aggregation
Consider replacing customers 1, … , q with mega-customer 0 by using variables {x(i, 0)} instead of variables {x(i, j) : j = 1, … , q}. We let D(0) denote the demand of mega-customer 0. Then, . Further, for j = 1, … , q, let θ(j) = D(j)/D(0). For all i, let c(i, 0) denote the cost of meeting unit demand at mega-customer zero from site i; we define c(i, 0) as: for all i. Also, let . Let P1C denote the condensed version of P1 corresponding to the
Computational results on customer aggregation for a real-world problem
We now report computational experiences on a real-world variant of the CFLP (refer Sankaran and Raghavan, 1997, for details of the problem). We present detailed results for six instances of the real-world problem.
We let PO and PC respectively denote the original and condensed versions of the real-world problem. Table 1 displays summary data for the two versions. Thus, the number of mega-customers is 100 and is about 22% of the total number of customers, which is 449.
The client company had two
Results on the suspension of VUB constraints
We now present results concerning the deterioration in the duality gap corresponding to the formulation P1 that results from dropping a subset of the VUB constraints. The results of this section are applicable to each node of the branch-and-bound tree, and not just to the root node. Hence, we let Ω and Ψ respectively denote the subsets of facilities that are open and closed at an arbitrary node in the tree; clearly, Ω and Ψ are disjoint. Further, we let P3 denote the subproblem corresponding to
Practical implications of the results
Both sets of results, namely those pertaining to customer aggregation and the suspension of VUB constraints, are relevant in any real-world instance of facility location problems in which cities and towns define ‘customers’ and their ‘demands’. Consider any metropolis and its (small) satellite townships. In general, the population of a metropolis exceeds (often, greatly) the cumulative demand of its satellite townships.
For such instances of P1, Proposition 1 and the ensuing heuristic for
Acknowledgment
The author is very grateful to an anonymous referee for informed comments on earlier versions that have helped to greatly improve the paper.
References (17)
Capacitated facility location: Separation algorithms and computational experience
Mathematical Programming
(1998)- et al.
Capacitated facility location: Valid inequalities and facets
Mathematics of Operations Research
(1995) - et al.
Erratum: Capacitated facility location: Valid inequalities and facets
Mathematics of Operations Research
(1996) - et al.
Polyhedral analysis and decompositions for capacitated plant location-type problems
Discrete Applied Mathematics
(1998) - Deng, Q., Simchi-Levi, D., 1993. Valid inequalities, facets, and computational experience for the capacitated...
- et al.
On worst-case aggregation analysis for network location problems
Annals of Operations Research
(1992) - et al.
Row-column aggregation for rectilinear distance p-median problems
Transportation Science
(1996) - Geoffrion, A.M., 1976. Customer aggregation in distribution modelling. Working paper No. 259, Western Management...
Cited by (12)
The continuous single source location problem with capacity and zone-dependent fixed cost: Models and solution approaches
2017, European Journal of Operational ResearchPrepositioning of supplies in preparation for a hurricane under potential destruction of prepositioned supplies
2013, Socio-Economic Planning SciencesCitation Excerpt :CPLEX performs satisfactorily even for relatively large problems. An approach that uses clusters of DP based on Sankaran [24] is presented to solve larger instances of the problem. A sensitivity analysis was conducted to test the impact of the parameters of our model on three different outcomes: total expected cost, number of selected SPs, and level of storage of the MDC.
End-of-line delivery vehicle routing optimization based on large-scale neighbourhood search algorithms considering customer-consumer delivery location preferences
2022, Advances in Production Engineering And ManagementLocation-routing problem of pickup point based on administrative address of customer points
2019, Jisuanji Jicheng Zhizao Xitong/Computer Integrated Manufacturing Systems, CIMSLocation planning and network design
2017, International Series in Operations Research and Management ScienceCoal logistics node location method based on analytic hierarchy process
2015, Nanjing Li Gong Daxue Xuebao/Journal of Nanjing University of Science and Technology