On A Stochastic Multi-Facility Location Problem

Publisher Website
Google Scholar
Abstract
This paper first discusses randomness in industrial multi-facility location problems. Different kinds of optimization criteria are then described for a stochastic location problem and the fractile approach is chosen. Seppälä's “Chance-Constrained Programming Algorithm” is used to solve the stochastic multi-facility location problems, where transportation costs are random variables and distances between facilities are Euclidean. The covariance matrix of the problem is defined by weighting two extreme cases, where in one case the random cost variables are totally correlated and in another case they are distributed independently of one another. Finally a numerical example is presented and solved. The solutions of deterministic and stochastic versions of multi-facility problems differ greatly from one another.