华南理工大学学报(自然科学版) ›› 2012, Vol. 40 ›› Issue (6): 126-131,158.

• 计算机科学与技术 • 上一篇    下一篇

基于多目标蚁群优化的知识即服务动态组合策略

贾瑞玉1 伍章俊2 张以文1,2   

  1. 1.安徽大学 计算机科学与技术学院,安徽 合肥 230039; 2.合肥工业大学 管理学院,安徽 合肥 230009
  • 收稿日期:2011-12-11 修回日期:2012-05-16 出版日期:2012-06-25 发布日期:2012-05-03
  • 通信作者: 贾瑞玉(1965-) ,女,副教授,主要从事智能软件、数据挖掘研究. E-mail:jiaruiyu267@ yahoo.com.cn
  • 作者简介:贾瑞玉(1965-) ,女,副教授,主要从事智能软件、数据挖掘研究.
  • 基金资助:

    国家"863”计划云制造主题项目( 2011AA040501) ; 国家自然科学基金资助项目( 70871033) ; 安徽省教育厅自然科学重点项目( KJ2011A006)

Dynamic KaaS Combination Strategy Based on Multi-Objective Ant Colony Optimization

Jia Rui-yuWu Zhang-junZhang Yi-wen1,2   

  1. 1. School of Computer Science and Technology,Anhui University,Hefei 230039,Anhui,China; 2. School of Management,Hefei University of Technology,Hefei 230009,Anhui,China
  • Received:2011-12-11 Revised:2012-05-16 Online:2012-06-25 Published:2012-05-03
  • Contact: 贾瑞玉(1965-) ,女,副教授,主要从事智能软件、数据挖掘研究. E-mail:jiaruiyu267@ yahoo.com.cn
  • About author:贾瑞玉(1965-) ,女,副教授,主要从事智能软件、数据挖掘研究.
  • Supported by:

    国家"863”计划云制造主题项目( 2011AA040501) ; 国家自然科学基金资助项目( 70871033) ; 安徽省教育厅自然科学重点项目( KJ2011A006)

摘要: 为了以Web 服务方式实现云计算环境下的知识共享和知识融合,提出了一种基于多目标蚁群优化的知识即服务组合策略.该策略中,结合云计算环境的动态性和知识即服务的质量规则,从知识服务提供者的角度构建了知识即服务动态组合模型; 同时,为了利用问题的特征信息引导蚂蚁的搜索行为,设计了蚁群算法相应的信息素和启发信息,从而实现多目标优化.在云计算平台下使用真实的Web 服务实例进行仿真实验,将该策略与基于遗传算法和协同进化算法的策略进行比较,结果表明,文中策略的性能和解的质量均明显较优.

关键词: 云计算, 多目标蚁群优化, 知识即服务

Abstract:

In order to implement the knowledge sharing and integration in the form of Web services in cloud computing environments,a KaaS ( Knowledge as a Service) combination strategy based on multi-objective ant colony optimization is proposed. In this strategy,a dynamic KaaS combination model, which takes into consideration the dynamic characteristics of cloud computing environments and the QoS ( Quality of Service) rules of KaaS,is established from the viewpoint of knowledge service provider. Then,by redesigning the corresponding pheromone and heuristic information of the ant colony algorithm,the features of the problem are used to guide the searching process,and the multi-objective optimization is thus achieved. Finally,a simulation is conducted with real Web services on the cloud computing platform. The results indicate that,as compared with the strategies based on the genetic algorithm and the coevolution algorithm,the proposed strategy is more effective in terms of performance and solution quality.

Key words: cloud computing, multi-objective ant colony optimization, knowledge as a service

中图分类号: