Show simple item record

dc.contributor.advisorHigle, Julia L.en_US
dc.contributor.authorHuang, Chih-yüan
dc.creatorHuang, Chih-yüanen_US
dc.date.accessioned2013-05-16T09:52:28Z
dc.date.available2013-05-16T09:52:28Z
dc.date.issued1987en_US
dc.identifier.urihttp://hdl.handle.net/10150/292045
dc.description.abstractWe show that, under certain conditions, instead of solving stochastic capacity expansion problems, we will obtain the same optimal solution by solving deterministic equivalent problems. Since only the first decision must be implemented immediately, knowing the optimal first decision is nearly as good as knowing the entire optimal sequences. Hence if we can solve the problem with 'big enough' finite horizon such that the first decision remains optimal for longer than this finite horizon, then we identify the 'big enough' finite horizon as forecast horizon. The forward dynamic programming recursion can be used to solve a finite horizon problem. An efficient forward algorithm has been developed to obtain the first optimal decision and forecast horizon. A heuristic algorithm also has been derived to prove an initial decision is within known error bound of the optimal first decision. Several examples are examined to investigate how a decision will be affected by randomness. (Abstract shortened with permission of author.)
dc.language.isoen_USen_US
dc.publisherThe University of Arizona.en_US
dc.rightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.en_US
dc.subjectIndustrial capacity -- Mathematical models.en_US
dc.titleAN ANALYSIS OF CAPICITY EXPANSION PROBLEMS WITH BACKORDERS AND STOCHASTIC DEMANDen_US
dc.typetexten_US
dc.typeThesis-Reproduction (electronic)en_US
dc.identifier.oclc18899660en_US
thesis.degree.grantorUniversity of Arizonaen_US
thesis.degree.levelmastersen_US
dc.identifier.proquest1332234en_US
thesis.degree.disciplineGraduate Collegeen_US
thesis.degree.disciplineSystems and Industrial Engineeringen_US
thesis.degree.nameM.S.en_US
dc.identifier.bibrecord.b16665442en_US
refterms.dateFOA2018-06-19T01:39:54Z
html.description.abstractWe show that, under certain conditions, instead of solving stochastic capacity expansion problems, we will obtain the same optimal solution by solving deterministic equivalent problems. Since only the first decision must be implemented immediately, knowing the optimal first decision is nearly as good as knowing the entire optimal sequences. Hence if we can solve the problem with 'big enough' finite horizon such that the first decision remains optimal for longer than this finite horizon, then we identify the 'big enough' finite horizon as forecast horizon. The forward dynamic programming recursion can be used to solve a finite horizon problem. An efficient forward algorithm has been developed to obtain the first optimal decision and forecast horizon. A heuristic algorithm also has been derived to prove an initial decision is within known error bound of the optimal first decision. Several examples are examined to investigate how a decision will be affected by randomness. (Abstract shortened with permission of author.)


Files in this item

Thumbnail
Name:
azu_td_1332234_sip1_m.pdf
Size:
1.417Mb
Format:
PDF

This item appears in the following Collection(s)

Show simple item record