Presumably, a facility would not be built at all unless it provided a service that some entity found "desirable." Conversely, even the most desirable of facilities also exhibit some undesirable characteristics. The fact is, most facilities cannot be classified as being either purely desirable or purely obnoxious. The location literature defines facilities falling between the two extremes as semi-desirable or semi-obnoxious and it is this type of facility that our research will address.; Chapter 2 builds a framework for modeling semi-obnoxious multifacility location problems. In that chapter we discuss three components of a facility location model---solution space, objectives, and interactions. Chapter 3 presents an algorithm that uses interval bisection to solve certain semi-obnoxious multifacility location problems. Chapter 4 discusses how set partitioning can be used to identify optimal solutions for a class of semi-obnoxious multifacility location problems wherein the facilities must be "pushed" as far away as possible from any protected points and customers have specified a minimum threshold distance from which to receive service. Chapter 5 develops a heuristic for a category of semi-obnoxious multifacility location problems that has the objective: maximize the minimum distance among the new facilities and any protected points. Finally, in Chapter 6 we summarize the main technical results of this dissertation and identify open problems for additional research.
展开▼