作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2007, Vol. 33 ›› Issue (13): 37-39. doi: 10.3969/j.issn.1000-3428.2007.13.013

• 博士论文 • 上一篇    下一篇

基于自适应算法的动态网格服务选择方法

李 清1,李志蜀1,朱明放1,殷 锋1,2,叶 军1,陈良银1   

  1. (1. 四川大学计算机学院,成都 610064;2. 西南民族大学计算机科学与技术学院,成都 610041)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-07-05 发布日期:2007-07-05

Approach of Dynamic Grid Service Selection Based on Self-adapting Algorithm

LI Qing1, LI Zhishu1, ZHU Mingfang1, YIN Feng1,2, YE Jun1, CHEN Liangyin1   

  1. (1. School of Computer, Sichuan Univ., Chengdu 610064; 2. School of Computer Science and Technology, Southwest University for Nationalities, Chengdu 610041)
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-05 Published:2007-07-05

摘要: 针对网格服务的动态性、时序性和随机性,给出了一种基于Q-learning的动态网格服务选择方法,用于求解具有不完全信息的网格环境中的服务组合。对满足马尔可夫决策过程的服务组合提出了一种支持不完备信息描述的网格服务描述模型,实现了对服务组合整个生命周期的描述。提出了一种改进的Q-learning 算法,动态、自适应地对服务选择中不同选择进行预估,并给出不同情况下的最优选择决策。仿真实验表明了该方法较传统的贪心选择算法具有优越性与实用性。

关键词: 网格服务组合, Q-learning, 马尔可夫决策过程

Abstract: In order to improve the efficiency of grid service selection, a new approach based on Q-learning is proposed. A new model based on Markov decision processes is proposed and the correlative novel algorithm is implemented with the adaptive ability of improved Q-learning for dynamic grid service selection. The experiment results show that the method is more effective than the traditional ones. Thus, it provides a good solution for grid service selection.

Key words: grid service composition, Q-learning, Markov decision processes(MDPs)

中图分类号: