Frank Ciarallo (Committee Member), Raymond Hill (Committee Member), Xinhui Zhang (Advisor)
Master of Science in Engineering (MSEgr)
Delivery operations use centralized warehouses to serve geographically distributed customers. Resources (e.g. personnel, trucks, stock, and equipment) are scheduled from the warehouses to distributed locations with the aim of: (a) meeting customer demands and, (b) rationalizing delivery operation costs. My thesis investigates the problem of clustering customers based on their geographical vicinity and their multi-period demands, while optimally scheduling resources. The problem addresses with-and-without capacity constraints of vehicles at the warehouse. This problem is proven to be NP-Hard. Hence, solutions using state-of-the-art exact methods such as branch and bound are not pertinent due to the computation complexity involved. We develop a K-means clustering algorithm for the initial solution and a tabu search heuristic that combines three advanced neighborhood search algorithms: (i) shift move, (ii) shift move with supernodes, and (iii) ejection chain with supernodes, to accelerate convergence.
Department or Program
Department of Biomedical, Industrial & Human Factors Engineering
Year Degree Awarded
Copyright 2008, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.