中文摘要：

在面向服务的环境下，单个Web服务往往不能满足用户的要求，这时就需将已有的单个Web服务进行组合，以便产生满足用户需求的、增值的组合服务．已有的服务组合方法都很少考虑Web服务的随机性和Internet环境的动态性，从而在服务选择过程中产生的规划都是静态规划，结果导致在服务组合时都以较大概率出现组合失败．针对上述问题，提出了Web服务各随机QoS指标的度量方法和自适应QoS管理体系结构，并利用随机型离散事件系统唯一的动态控制方法——马尔可夫决策过程（MDP），设计出随机QoS感知的可靠Web服务组合算法．实验结果表明，考虑随机性的QoS度量方法和QoS管理体系结构，以及平衡了“风险”与“报酬”的MDP有效地提高了服务组合成功率．

英文摘要：

In the service-oriented environment, a single Web service can hardly satisfy the given request, so the composition of multiple Web services is required to fulfill the goal. Without considering the inherent stochastic and dynamic nature of Web service, the existing composition methods mostly generate static plans. As a result, Web service composition often terminates with failure inevitably. In this paper, metrical methods of several random QoS dimensions and QoS Management Architecture are presented, and one reliable Web service composition algorithm is also designed based on markov decision process （MDP）-only dynamic controlling method of stochastic discrete event system （SDES）. Experimental results demonstrate the success rate of Web service composition has been improved greatly.