A heuristic algorithm for Facility location and demand allocation In presence of barriers in facilities network

In this study single resource capacitated Multi-Facility Weber Problem (SSCMFWP) is considered. This problem known as continuous location–allocation problem, has a non-convex objective function and when in network facilities there is some barrier to locate facilities or transfer through network, has non-convex solution space as well. We consider two kinds of barriers, namely Forbidden regions and Barriers to travel. The first are barriers that a facility couldn’t be located at them but traveling through them is possible (for example a lake). The second are barriers that we can neither locate a facility at them nor travelling through them (for example a mountain).In this paper for forbidden barrier we extend the algorithm suggested by Manzour-al-Ajdad et al (Manzour,2012) to the case of forbidden barriers. To tackle with barrier to travel we use visibility concept in graph representation of network and Dijkstra algorithm to find shortest path between facilities as their distances.The heuristic suggested here developed in two phases. In the first phase, the initial locations of new facilities are determined. In phase II the allocation of demand points is done and this phase is repeated until the algorithm reached terminations condition.

کلیدواژه ها:

Capacitated location –allocation ، Forbidden regions ، barrier to travel ، visibility concept ، heuristic algorithm