Project Description

This dissertation develops a set of optimization models addressing integrated facility location in the framework of regional development plans. In this context, rather than defining whether a particular facility should be located in a particular urban center in some region (as is usual in location modeling), the aim is to redefine the region’s urban hierarchy – that is, the level of hierarchy to assign to the urban centers of the region, each level being characterized by a class of facilities.

The objective of the models is to maximize the accessibility of population to the different classes of services available in the region (that is, to minimize the total travel distance/time between the population centers and the facilities where the services are provided). While the initial location models assume the transportation network of the region as given, the final models are also aimed at determining which links of the transportation network should be improved. The models explicitly represent the influence of the decisions about urban hierarchy on the spatial distribution of population in future periods (that is, on the future demand for service). The relationship between population dynamics and urban hierarchy is confirmed through a statistical study – a spatial error model is shown to describe accurately such relationship for the case of Portugal during the period 1991-2001.

The dissertation is also concerned with model solution. As a starting point, the mixed-integer models are solved through a general-purpose optimization package. In general, this requires a large computing effort. Therefore, heuristic methods are developed. These methods are also used to solve the non-linear models. The heuristic methods developed (in particular, tabu search and variable neighborhood search) provide optimum or near-optimum solutions while requiring a much smaller computer effort.

Research Team

CITTA

• João Bigotte
• António Pais Antunes (supervisor)

University of Toronto

• Oded Berman

Financial Support

• FCT

Stage of Progress

• Concluded in 2009