Optimal budget allocation for discrete-event simulation experiments
Optimal budget allocation for discrete-event simulation experiments
复制标题
DOI：
10.1080/07408170903116360
复制
发表时间：
2010-01-01
期刊：
影响因子：
--
通讯作者：
Chen, Hsiao-Chang
中科院分区：
文献类型：
--
作者：
Chen, Chun-Hung;Yuecesan, Enver;Chen, Hsiao-Chang
仿真在分析离散事件系统中起着至关重要的作用，特别是在比较不同的系统设计以优化系统性能方面。然而，使用仿真来分析复杂系统可能既昂贵又耗时。提出了一种有效的离散事件仿真实验计算预算智能分配算法。这些算法动态地确定所有仿真实验的仿真长度，从而在给定计算预算的约束下显著地提高了仿真效率。通过数值实验，将该算法与传统的两阶段排序选择算法进行了比较。虽然所提出的方法是基于几何学，数值结果表明，它是更有效的比比较程序。
Simulation plays a vital role in analyzing discrete-event systems, particularly in comparing alternative system designs with a view to optimizing system performance. Using simulation to analyze complex systems, however, can be both prohibitively expensive and time-consuming. Effective algorithms to allocate intelligently a computing budget for discrete-event simulation experiments are presented in this paper. These algorithms dynamically determine the simulation lengths for all simulation experiments and thus significantly improve simulation efficiency under the constraint of a given computing budget. Numerical illustrations are provided and the algorithms are compared with traditional two-stage ranking-and-selection procedures through numerical experiments. Although the proposed approach is based on heuristics, the numerical results indicate that it is much more efficient than the compared procedures.