Model-driven multi-view conceptual architecture design in a PHM system environment

摘要

Today's systems or devices in the Internet of Things (IoT) applications are more and more complex. Analytic facilities are typically deployed in these systems for Prognostics and Health Management (PHM) purposes. One of the main challenges for architecting such control and diagnostics systems is their restrictive performance requirements, which directly impact the architecture and subsequent detailed designs. This paper presents a case study on the conceptual architecture design in such an environment, and co-develops the architecture and the domain solution to improve prototyping rapidness. First, on applying systems thinking and technologies, during architecting, we include the tasks of meeting both functional and performance requirements especially with respect to analytics, in a single architecture description framework. Second, we study a domain-specific issue, where un-attempted methods must be researched, and provide a solution that re-uses existing system elements, through which the mutual impacts between the information from both architecture and domain solutions are shown. Practical and important issues including benefits, standards, and implementations are also explored as part of the required tasks for architecting. Results including architecture designs and empirical performance research are illustrated, to indicate a systematic design process for systems in a PHM environment.