为提高调控云体系下在线计算软件服务化水平,分析了调控云环境下在线计算软件的服务场景,设计了在线计算软件服务的层次结构。为适应这种"痩客户端、胖服务器端"的类B/S(Browser/Server)软件服务访问模式,将在线计算流程进行...为提高调控云体系下在线计算软件服务化水平,分析了调控云环境下在线计算软件的服务场景,设计了在线计算软件服务的层次结构。为适应这种"痩客户端、胖服务器端"的类B/S(Browser/Server)软件服务访问模式,将在线计算流程进行分析拆解封装成计算服务单元。采用多代理机制设计了层次化的在线计算服务框架,以搭积木组合排列的方式提供在线计算软件服务。最后在JADE(Java Agent Development Framework)多代理框架下设计实现了调度员潮流计算软件服务,进行了相关测试与分析。并对提供潮流计算软件服务的同时所引起的效率损失问题提出了进一步优化的技术措施。展开更多
Based on the immune mechanics and multi-agent technology, a multi-agent artificial immune network (Maopt-aiNet) algorithm is introduced. Maopt-aiNet makes use of the agent ability of sensing and acting to overcome pre...Based on the immune mechanics and multi-agent technology, a multi-agent artificial immune network (Maopt-aiNet) algorithm is introduced. Maopt-aiNet makes use of the agent ability of sensing and acting to overcome premature problem, and combines the global and local search in the searching process. The performance of the proposed method is examined with 6 benchmark problems and compared with other well-known intelligent algorithms. The experiments show that Maopt-aiNet outperforms the other algorithms in these benchmark functions. Furthermore, Maopt-aiNet is applied to determine the Murphree efficiency of distillation column and satisfactory results are obtained.展开更多
文摘为提高调控云体系下在线计算软件服务化水平,分析了调控云环境下在线计算软件的服务场景,设计了在线计算软件服务的层次结构。为适应这种"痩客户端、胖服务器端"的类B/S(Browser/Server)软件服务访问模式,将在线计算流程进行分析拆解封装成计算服务单元。采用多代理机制设计了层次化的在线计算服务框架,以搭积木组合排列的方式提供在线计算软件服务。最后在JADE(Java Agent Development Framework)多代理框架下设计实现了调度员潮流计算软件服务,进行了相关测试与分析。并对提供潮流计算软件服务的同时所引起的效率损失问题提出了进一步优化的技术措施。
基金Supported by the National Natural Science Foundation of China (61271137)Public Science and Technology Research Funds Projects of Zhejiang Province (2011C21077)the Natural Science Foundation of Ningbo City (2011A610173)
文摘Based on the immune mechanics and multi-agent technology, a multi-agent artificial immune network (Maopt-aiNet) algorithm is introduced. Maopt-aiNet makes use of the agent ability of sensing and acting to overcome premature problem, and combines the global and local search in the searching process. The performance of the proposed method is examined with 6 benchmark problems and compared with other well-known intelligent algorithms. The experiments show that Maopt-aiNet outperforms the other algorithms in these benchmark functions. Furthermore, Maopt-aiNet is applied to determine the Murphree efficiency of distillation column and satisfactory results are obtained.
