We consider a selection of problems from civil infrastructure network design that are of great importance in modern urban planning but have,... Show moreWe consider a selection of problems from civil infrastructure network design that are of great importance in modern urban planning but have, until relatively recently, gone largely ignored in mathematical literature. Each of these problems is approached from the perspective of network optimization-based modeling, with a major focus placed on the development of efficient solution algorithms.We begin with a study of the phenomenon of interdependent civil infrastructure networks, wherein the functionality of one network (such as a telecommunications system) requires the input of resources from another network (such as the electrical power grid). We first consider a linear relaxation of an established binary interdependence minimum-cost network flows model, including its unique modeling applications and its use as part of a randomized rounding approximation algorithm for the mixed integer model. We also develop a generalized network simplex algorithm for the efficient solution of this generalized minimum-cost network flows problem. We then move on to consider a trilevel network interdiction game for use in planning the fortification of interdependent networks subject to targeted attacks. A variety of solution algorithms are developed for both the binary and the linear interdependence models, and the linear interdependence model is used to develop an approximation algorithm for the more computationally expensive binary model.We then develop a public transit network design model which incorporates a social access objective in addition to traditional operator cost and user cost objectives. The model is meant for use in planning minor modifications to a public transit network capable of improving equity of access to important services while guaranteeing that service levels remain within a specified tolerance of their initial values. A hybrid tabu search/simulated annealing algorithm is developed to solve this model, which is then applied to a test case based on the Chicago public transit network with the objective of improving equity of primary health care access across the city. Show less
Query
(-) mods_name_creator_namePart_mt:"Rumpf, Adam Andrew"