鉴于物联网资源的特性,现有的应用方式难以有效地使用和管理物联网中的资源.这种现状阻碍了物联网的发展和实际应用.其中,如何自动接入异构资源,如何支持资源的演化和保证物联网服务的连续性,如何整合来自异构资源的信息,如何将原始信息转化为有意义的情境信息都是急需解决的问题.文中针对上述问题,结合物联网资源特性,提出一种支持资源动态演化和反演的资源模型和资源框架.资源模型采用本体方式,对现有W3CSSN等本体进行了扩展,使用LinkedData技术,使之能够表达领域知识和基于事件的反演.框架基于SOA架构,以资源模型为中心,采用资源实体绑定并结合领域知识自动解释信息,采用事件树和诱导演绎推理自动产生事件,并基于资源反演机制解释历史数据和事件.案例分析中,煤矿综合监控顸警系统对模型和框架进行了验证,证明其能够较为有效的解决上述问题.
In view of the characteristics of IoT resources, the current architectures could not effectively use and manage IoT resources. Such a situation hindered IoT development and practical application. How to automatically access heterogeneous resources, how to sup- port the evolution of resources and maintain service continuity, how to integrate the data which comes from heterogeneous sources, how to make raw data meaningful to applications and naive users are some of the key challenges. To address these issues, this paper proposes a resource model and resource framework supporting dynamic evolution and devolution. The resource model uses of Linked Data and ex- tends the existing ontologies, such as W3C SSN, etc. This resource model can express domain knowledge and support event-based de- volution. SOA-based framework uses resource-entity binding and domain knowledge to automatic interpret information, uses event tree and PCT algorithm to generate events, interprets history data and events based on resource devolution. The validation of the resource model and framework is shown through the CCMWS ( Coal mine comprehensive monitoring and early warning system ).