On solving complex multi-period location models using simulated annealing

Abstract

This paper describes a study aimed at evaluating the capabilities of simulated annealing in dealing with complex, real-world multi-period location problems raised by school network planning in Portugal. The problems were formulated as mixed-integer linear optimization models. The models allow for facility closure or size reduction besides facility opening and size expansion, with sizes possibly limited to a set of pre-defined standards. They assume facility costs to be divided into a fix component and two variable components, respectively dependent on facility size and facility attendance. Results obtained through the study indicate that simulated annealing can be a useful tool for solving these kinds of models.