Optimal way to find the optimal design for discrete-event stimulation experiments

Chun-Hung Chen; Liyi Dai; Enver Yucesan

doi:10.1117/12.319339

24 August 1998 Optimal way to find the optimal design for discrete-event stimulation experiments

Chun-Hung Chen, Liyi Dai, Enver Yucesan

Proceedings Volume 3369, Enabling Technology for Simulation Science II; (1998) https://doi.org/10.1117/12.319339
Event: Aerospace/Defense Sensing and Controls, 1998, Orlando, FL, United States

Abstract

Simulation plays a vital role in analyzing many discrete- event systems, particularly, in comparing alternative system designs with a view to optimize system performance. Usually, using simulation to analyze complex systems can be both prohibitively expensive and time consuming. We present an effective algorithm to intelligently allocate computing budget for discrete-event simulation experiments. These algorithms dynamically determine the best simulation lengths for all simulation experiments and thus significantly reduce the total computation cost for a desired confidence level. This provides an optimal way to find an optimal design. We also compare our algorithms with traditional two-stage procedures and with the techniques for the multi-armed bandit problem through numerical experiments. Numerical testing shows that our approach is more than fifteen times faster than the compared methods for the same simulation quality requirements.

Citation Download Citation

Chun-Hung Chen, Liyi Dai, and Enver Yucesan "Optimal way to find the optimal design for discrete-event stimulation experiments", Proc. SPIE 3369, Enabling Technology for Simulation Science II, (24 August 1998); https://doi.org/10.1117/12.319339

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available