Efficient selection of a set of good enough designs with complexity preference

This thesis briefly reviews the important methods involved in solving the best design selection problem in the discrete-event system simulation. The selection of one or several best designs is a common problem people meet in real situations. The research originally focused on the one best design selection problem, the two-stage procedure. Later, there was literature about multiple designs selection problems which are useful in the global optimization as well. Most recently, some researchers have studied the problem of selecting one simplest sufficiently good design applicable to the node activation rule in the wireless sensor networks. However, the problem for selecting several simplest good enough designs is still open for consideration. The second part of the thesis introduces two new algorithms for solving the selection problem related to the designs mentioned above. These two algorithms OCBA-mSG and OCBA-bSG allocate the simulation budget efficiently to identify a subset of m simplest and good enough designs among a total of K (K > m) designs. The numerical results show that both OCBA-mSG and OCBA-bSG outperform some other approaches on the test problems.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Industrial Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2011

Author dc:creator

• Yan, Shen

Contributors dc:contributor

• Zhou, Enlu

• discrete-event system simulation
• best design selection problem
• Optimal Computing Budget Allocation

Statement dc:rights

• Copyright 2011 Shen Yan

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/26270

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/26270