Simulation II
Session: SC34
Date/Time: Sunday 13:00-14:30
Type: Contributed
Sponsor:
Track:
Cluster:
Room: Rm. 607
Chair: Paul Savory
Chair Address: University of Nebraska, 175 Nebraska Hall, Lincoln, NE 68588-0518
Chair E-mail: psavory@unl.edu, http://www.engr.unl.edu/savory/
Chair:
Chair Address:
Chair E-mail:
- SC34.1 Using Simulation for Finding Optimal Hedging Points of Production Controllers
- Gul Gurkan;
Tilburg University, Ctr. for Economic Res., PO Box 90153, Tilburg, 5000 LE , The Netherlands;
ggurkan@kub.nl
- Fikri Karaesmen;
Universite Pierre et Marie Curie, Lab. d'Info. de Paris 6, 4 Place Jussieu, Paris, 75252 , France;
fikri.karaesmen@lip6.fr
We model the dynamics of a fluid production controller using generalized semi-Markov processes and compute sensitivity estimates of cost functions, consisting of inventory and backlog costs, with respect to hedging points. We then discuss how they are used together with sample-path optimization for finding optimal hedging points. - SC34.2 Optimal Kanban Allocation in JIT Production Systems Using Neural Networks
Simulation has been used to evaluate kanban allocations in JIT systems. However, it is not practical to do an exhaustive search by simulation since the number of possibilities is very large. We train a neural network by simulation results and utilize it to determine the optimum kanban allocation among all possibilities. - SC34.3 Quantile Estimation in Steady-State Simulation Using Bonferroni & Bootstrap Methods
- Yoon G. Hong;
Hansung University, Dept. IE, 39 2-Ka Samsun-Dong, Sungbuk-Ku, Seoul, 136-792 , Korea;
yhong@hsel.hansung.ac.kr
- Se Y. Kim;
Korea University, Dept. of IE, 1, 5-ka, Anam-Dong, Sungbuk-ku, Seoul, 136-701 , Korea;
- Seong I. Kim;
Korea University, Dept. of IE, 1, 5-ka, Anam-Dong, Sungbuk-ku, Seoul, 136-701 , Korea;
tennis@kuccnx.korea.ac.kr
Both bootstrap and Bonferroni simultaneous approach are combined to estimate the quantile in steady-state simulation. Simple queueing system and time series models are employed to see the model's fidelity. Comparisons are made and results show that the narrower interval is obtained with an acceptable confidence level. - SC34.4 withdrawn author request
For information on individual presentations, please contact the authors
directly.
Return to Conference home page
|
