Nowadays, the emergence of large-scale and highly distributed cyber-physical systems (CPSs) in applications including Internet of things (IoT), cloud computing, mobility, Big Data, and sensor networks involves that architecture models have to work in an open and highly dynamic world. This fact increasingly highlights the importance of designing real-time and intelligent CPSs that are capable of decision making. In this paper, an Intelligent Architecture model for CPS management (IA-CPS) based on online processing is proposed for this purpose. Specifically, it can manage simple and complex events based on a service-oriented architecture and directed by an event-driven architecture, using event processing technology as Complex Event Processing (CEP). The novelty of our approach relies on the fact that the proposed architecture is service-oriented, which models the functionalities of the event-driven system. This gives us the possibility to offer a flexible service catalog, allowing us to connect to the system on any kind of device and interact in different scenarios. The model has been applied to two use cases: processing images from video surveillance cameras, and processing of consumption data captured by water and energy sensors installed in end-user environments.

如今，在物联网 (IoT)、云计算、移动性、大数据和传感器网络等应用中，大规模和高度分布式的网络物理系统 (CPS) 的出现涉及架构模型必须在开放和开放的环境中工作。高度动态的世界。这一事实越来越突出了设计能够做出决策的实时和智能 CPS 的重要性。为此，本文提出了一种基于在线处理的 CPS 管理智能架构模型（IA-CPS）。具体来说，它可以基于面向服务的架构，以事件驱动架构为导向，使用复杂事件处理（CEP）等事件处理技术来管理简单和复杂的事件。我们方法的新颖之处在于所提出的架构是面向服务的，它对事件驱动系统的功能进行建模。这使我们能够提供灵活的服务目录，使我们能够在任何类型的设备上连接到系统并在不同的场景中进行交互。该模型已应用于两个用例：处理来自视频监控摄像头的图像，以及处理安装在最终用户环境中的水和能源传感器捕获的消耗数据。