Contatto di riferimento: Prof. Daniele Vigo
Abstract
In the last ten years the research activity on location-routing problem (LRP) has significantly increased and many effective exact and approximated approaches have been presented for its solution. Following from this, great attention has been also devoted to several variants of the LRP, most of them coming from real applications. In particular, with reference to the city logistic (CL) context, several variants of the LRP have been proposed to design single-tier and two-tier freight distribution systems. In this case, LRP location decisions concern the number and position of one or more types of logistic facilities, whereas, routing decisions, concern the definition of dedicated or multi-stop routes of homogeneous or heterogeneous vehicle fleets.
However most of the present literature deals with single-commodity LRP, whereas the multi-commodity case is still scarcely investigated. This issue can assume an important role in LRP real applications and in particular in the CL context. Indeed, the freight to be distributed in an urban area are highly heterogeneous and consequently the problem is inherently multi-commodity.
In order to contribute to fill this gap, an original formulation for the multi-commodity LRP is presented. The problem is referred to as the flow intercepting facility location-routing problem (FIFLOR), since its formulation is obtained merging the flow intercepting facility location and the vehicle routing models. A branch-and-cut algorithm for its solution is described and results obtained on test instances reproducing different urban network topologies are discussed.