Dynamic multi-level capacitated and uncapacitated location problems: an approach using primal-dual heuristics

In this paper several dynamic multi-level location problems are formulated as mixed-integer linear programs. Both uncapacitated and capacitated versions of the problem are studied. The models presented are more complete than the ones known from the literature: they are dynamic and consider the possibility of a facility being opened, closed and reopened more than once during the planning horizon. They may include both upper and lower limits on the used capacity of each facility and may also consider the situation where there is no flow conservation in the intermediate facilities. Primal-dual heuristics were developed to solve the proposed models, having as main objective the capability of finding good primal solutions in reasonable computational times. Computational results are presented and discussed.