摘要
为实现制造网格应用的自动化,提出了一个基于领域本体(ontology)的服务自动组合体系结构及其相应的算法。领域本体基于TOVE、STEP、PSL三个成熟的制造本体和网格服务概念模型构建。体系结构以语义结构良好的用户目标作为输入,输出预约的可执行工作流;通过基于领域本体推理的反向递归组合算法,能够实现本地仓库中工作流的重用和网格范围内服务的新工作流组合;支持抽象、具体两个层面的松耦合工作流组合(对应于编排(Choreography)和编制(Orchestration));支持面向服务质量的工作流优化选择和工作流预约。实现了一个原型系统,实例的实验结果证明了该体系结构和算法的有效性。
In order to realize automatic applications of Manufacturing Grid （ MG）, a framework along with its algorithms for automatic service composition, were proposed based on domain-speCified ontology. The domain-specific ontology was constructed on three popular manufacturing ontologies, including TOVE, STEP, PSL, and was adapted to the concept model of MG service （service ontology）. The framework accepted well-formed user goals as inputs and produced reserved executable workflows. It could reuse locally stored workflows processed previously, and compose new workflows from MG services. Two layers of loosely-coupled workflow composition （corresponding to Choreography and Orchestration respectively） were supported. Optimal workflows were selected by the criterion of user-specified Quality of Service （QoS）. A prototype system was implemented. Experiments on an example of airplane part show that the framework along with its algorithms has good performance.
出处
《计算机应用》
CSCD
北大核心
2006年第1期57-60,共4页
journal of Computer Applications
关键词
制造网格
语义网格
服务自动组合
本体
语义网
manufacturing grid
semantic grid
automatic service composition
ontology
semantic Web
