This paper demonstrates the flexible methodology of modelling cyber-physical systems (CPSs) using the INTO-CPS technology through co-simulation based on Functional Mock-up Units (FMUs). It explores a novel method with two main co-simulation phases: homogeneous and heterogeneous. In the first phase, high-level, abstract FMUs are produced for all subsystems using a single discrete-event formalism (the VDM-RT language and Overture tool). This approach permits early co-simulation of system-level behaviours and serves as a basis for dialogue between subsystem teams and agreement on interfaces. During the second phase, model refinements of subsystems are gradually introduced, using various simulation tools capable of exporting FMUs. This heterogeneous phase permits high-fidelity models of all subsystems to be produced in appropriate formalisms. This paper describes the use of this methodology to develop a USB stick production line, representing a smart system of systems. The experiments are performed under the assumption that the orders are received in a Gaussian or Uniform distribution. The focus is on the homogeneous co-simulation phase, for which the method demonstrates two important roles: first, the homogeneous phase identifies the right interaction protocols (signals) among the various subsystems, and second, the conceptual (system-level) parameters identified before the heterogeneous co-simulation phase reduce the huge size of the design space and create stable constraints, later reflected in the physical implementation.

中文翻译：

---

通过协作建模对网络物理生产系统进行早期分析

本文演示了通过基于功能样机（FMU）的协同仿真，使用INTO-CPS技术对网络物理系统（CPS）建模的灵活方法。它探索了一种具有两个主要协同仿真阶段的新方法：同构和异构。在第一阶段，使用单个离散事件形式主义（VDM-RT语言和Overture工具）为所有子系统生成高级抽象FMU。这种方法允许早期对系统级行为进行共同仿真，并作为子系统团队之间对话和接口协议的基础。在第二阶段中，使用各种能够导出FMU的仿真工具逐步引入子系统的模型优化。这个异构阶段允许以适当的形式主义生成所有子系统的高保真模型。本文介绍了使用这种方法开发USB记忆棒生产线的方法，该生产线代表了系统的智能系统。实验是在假设订单以高斯或统一分布接收的前提下进行的。重点放在同构协同仿真阶段，该方法展示了两个重要作用：首先，同构阶段标识各个子系统之间的正确交互协议（信号），其次，确定概念（系统级）参数在异构协同仿真阶段之前，减小了设计空间的巨大规模并创建了稳定的约束，后来反映在物理实现中。实验是在假设订单以高斯或统一分布接收的前提下进行的。重点放在同构协同仿真阶段，该方法展示了两个重要作用：首先，同构阶段标识各个子系统之间的正确交互协议（信号），其次，确定概念（系统级）参数在异构协同仿真阶段之前，减小了设计空间的巨大规模并创建了稳定的约束，后来反映在物理实现中。实验是在假设订单以高斯或统一分布接收的前提下进行的。重点是同构协同仿真阶段，该方法展示了两个重要作用：首先，同构阶段标识各个子系统之间正确的交互协议（信号），其次，确定概念（系统级）参数在异构协同仿真阶段之前，减小了设计空间的巨大规模并创建了稳定的约束，后来反映在物理实现中。