Heuristic Approaches to a Sequential Variant of the p-Supplier Problem

A sequential location problem of facilities on a network is considered, which is viewed as a variant of the p-supplier problem. The traditional p-supplier problem asks to find a subset with cardinality p of facility location candidates on a given network which minimizes the maximum of the distance from a client to the nearest facility, while the variant to be discussed in this paper also asks to find an opening sequence of the p facilities in order to minimize the total of the weighted distance sums over all clients during the constructing horizon. An iterative improvement algorithm based on the simulated annealing is designed for the sequential location problem, where a swapping neighborhood structure in a list of all facility location candidates is explored. Numerical experiments are conducted to demonstrate the improvement obtained by the proposed simulated annealing algorithm.