Authors:
Gabriela Iriarte
;
Pablo Escalona
;
Alejandro Angulo
and
Raul Stegmaier
Affiliation:
Universidad Técnica Federico Santa María, Chile
Keyword(s):
Weber Problem, Fixed Costs, Delaunay Triangulation, Kriging Interpolation.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Artificial Intelligence
;
e-Business
;
Enterprise Information Systems
;
Industrial Engineering
;
Knowledge Discovery and Information Retrieval
;
Knowledge-Based Systems
;
Logistics
;
Mathematical Modeling
;
Methodologies and Technologies
;
Operational Research
;
Optimization
;
Supply Chain Management
;
Symbolic Systems
Abstract:
This paper analyzes the location of a distribution center in an urban area using a single-source Weber problem with continuous piecewise fixed cost to find a global optimal location. The fixed cost is characterized by a Kriging interpolation method. To make the fixed cost tractable, we approximate this interpolation with a continuous piecewise function that is convex in each piece, using Delaunay triangulation. We present a decomposition formulation, a decomposition conic formulation and a conic logarithmic disaggregated convex combination model to optimally solve the single-source Weber problem with continuous piecewise fixed cost. Although our continuous approach does not guarantee the global optimal feasible location, it allows us to delimit a zone where we can intensify the search of feasible points. For instances we tested, computational results show that our continuous approach found better locations than the discrete approach in 23.25% of the instances and that the decomposi
tion formulation is the best one, in terms of CPU time.
(More)