Detail Special models for a multi-item, multi-location, multi-echelon, multi-period inventory system with centralized replenishments and restricted transshipments Bibliografi Author: Bauskar, Milind Ekanath ; Schmidt, Charles P. (Advisor) Topik: OPERATIONS RESEARCH|ENGINEERING; INDUSTRIAL Bahasa: (EN )    ISBN: 0-591-73244-0     Penerbit: University of Alabama Press     Tahun Terbit: 1997     Jenis: Theses - Dissertation Fulltext: 9821522.pdf (0.0B; 20 download) Abstract The control and maintenance of inventory is a strategic problem for a wholesale/distribution business. Multi-echelon inventory deals with inventory problems comprising multiple interconnected production and/or distribution entities. A modernistic multi-echelon, multi-item model presented in this work taps into the advances in information technologies and emphasis on supply chain management and employs the concept that centralization of multi-location inventories leads to statistical economies of scale. A comprehensive multi-item, multi-echelon model called Complete Plan is developed that has two components: Allocation Plan and Transshipment Plan. Allocation Plan dynamically allocates system orders amongst the locations, based on equal fractile allocations. Transshipment Plan is employed every working day to arrange the restricted transshipments from the central location. Transshipment Plan employs a heuristic to solve complex vehicle-routing problems. Since the mathematical modeling of the problem is extremely complex, the tool of simulation is employed to design and optimize Complete Plan models. Periodic review (R, s, S) policies are order-up-to policies where one would order up-to S whenever the inventory position falls to or below s at the time of review. The models assume deterministic lead time, non-perishable commodity, stochastic demand, and full backlogging. The decentralized models in this research use near optimal (R, s, S) policies at individual locations whereas the Complete Plan models--in a centralized inventory system--are simulation-optimized using a kind of neighborhood search. A systematic design and analysis of experiments is performed mainly to study whether the selection of model has any impact on the total annual costs of an inventory system. The discovery is that indeed application of Complete Plan models, instead of decentralized mode of operation, would reduce the overall costs of an inventory system. These gains are magnified even further when the demand has high mean and high variability. Of course, the factor of transportation holds sway over whether cp models would fare better than dc models. Intuitive results are discussed and future directions suggested. Opini Anda Klik untuk menuliskan opini Anda tentang koleksi ini! Lihat Sejarah Pengadaan  Konversi Metadata   Kembali

Copyright © 2006, 2007 Unika Atma Jaya, all rights reserved