| 受控M/G/1排队系统的性能优化及迭代算法 |
| |
| 引用本文： | 代桂平,殷保群,王肖龙,奚宏生.受控M/G/1排队系统的性能优化及迭代算法[J].系统仿真学报,2004,16(8):1683-1685. |
| |
| 作者姓名： | 代桂平 殷保群 王肖龙 奚宏生 |
| |
| 作者单位： | 中国科学技术大学自动化系,合肥,230027 |
| |
| 基金项目： | 国家自然科学基金(60274012)，安徽省自然科学基金(01042308) |
| |
| 摘 要： | 通过嵌入Markov链的方法,讨论了受控M/G/1排队系统,在无限水平平均代价准则下的最优平稳策略问题。定义了平均Poisson方程,导出了平均代价模型在紧致行动集上的最优性方程。最后给出了一个求解最优平稳策略的迭代算法,并给出了一个仿真实例来说明该算法的应用。
|
| 关 键 词： | 受控M/G/1排队系统 平均代价准则 紧致行动集 最优平稳策略 迭代算法 |
| 文章编号： | 1004-731X(2004)08-1683-03 |
| 修稿时间： | 2003年6月24日 |
Performance Optimization and Iteration Algorithm for Controlled M/G/1 Queuing Systems |
| |
| DAI Gui-ping,YIN Bao-qun,WANG Xiao-long,XI Hong-sheng.Performance Optimization and Iteration Algorithm for Controlled M/G/1 Queuing Systems[J].Journal of System Simulation,2004,16(8):1683-1685. |
| |
| Authors: | DAI Gui-ping YIN Bao-qun WANG Xiao-long XI Hong-sheng |
| |
| Abstract: | By the embedded Markov chain, the problems of optimal stationary policies are studied for controlled M/G/1 queuing systems with the infinite horizon average-cost criteria. The average-Poisson equation is defined for infinite horizon average model and the optimality equation is induced in a compact action set. In the end, an iteration algorithm to find the optimal stationary policy is suggested and a numerical example is proposed to illustrate the application of the algorithm. |
| |
| Keywords: | controlled M/G/1 queuing systems average-cost criteria compact action set optimal stationary policy iteration algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
