|
Graduation Thesis from 2010
Ahn, Hyoung Dae
M.S. |
A Dynamic Lot Sizing Model for Remanufacturing Systems This paper considers the problem of determining the lot sizes that satisfy the time-varying demands of a single remanufactured product over a given planning horizon Remanufacturing, in which used or end-of-life products are restored to like-new condition, has three main processes, disassembly, reprocess and reassembly, and hence the lot sizes are determined for each of the three processes. The objective is to minimize the sum of setup and inventory holding costs at the three processes over the planning horizon. To represent the problem mathematically, we suggest an integer programming model that combines the existing models for disassembly and assembly systems. Then, due to the complexity of the problem, we suggest two dynamic programming based heuristic algorithms, aggregated and separated ones, after characterizing the properties of the problem. Computational experiments were done on a number of test instances, and the results are re-ported. |
Ha, Kyoung Woon
M.S. |
Priority rule based heuristics for the team orienteering problem Team orienteering, an extension of single-competitor orienteering, is the problem of determining multiple paths from a starting point to a finishing point for a given allowed time or distance limit fixed for each of the paths while maximizing the total collected score. Each path is through a subset of nodes, each of which has an associated score. The team orienteering problem has many applications such as home fuel delivery, college football players recruiting, service technicians scheduling, military operations, etc. Unlike the previous optimal and heuristic algorithms that are not suitable for practically very large-sized instances, we suggest two types of fast heuristics ? serial and parallel ones ? in which all nodes are listed in an order using a priority rule and then the paths are constructed according to this order. To show the performances of the fast heuristics, computational experiments were done on the small-to-medium sized benchmark instances up to 100 nodes and randomly generated large-sized test instances up to 1000 nodes, and the results show that some of the heuristics give reasonable quality solutions within very short computation time. |
Shim, Ik Soo
M.S. |
A study on Capacitated Lot-sizing and Scheduling with sequence dependent setups This paper considers the problem of determining the lot sizes and the sequence of lots while satisfying the demand requirement and the machine capacity over a planning horizon, called the capacitated lot-sizing and scheduling problem (CLSP) in the literature, for the objective of minimizing the sum of setup and inventory holding costs. In particular, we consider sequence-dependent setup costs that depend on the type of lot just completed and on the lot to be processed, and setup state preservations in which the setup state at one period can be carried over to the next period. Due to the complexity of the problem, we suggest two-stage heuristics in which an initial solution is obtained and then it is improved using a multi-pass interchange method. The heuristics are compared with the best existing heuristic in two ways: (a) real instances obtained from a case study on a paper remanufacturing system that produces several types of corrugated cardboards; and (b) various randomly generated test instances, and the results show that the two-stage heuristics suggested in this paper outperform the existing one. |
|
