HybridSim: A Modeling and Co-simulation Toolchain for Cyber-Physical Systems
Baras, John, S.
Date: October 30 - November 01, 2013
Cyber-physical systems (CPS) involve communication networks, computation algorithms, control systems and physical systems. Many CPS, such as Smart Buildings, are subject to very expensive deployment costs and complex network interactions. Thus comprehensive modeling and simulation of such systems are crucial to ensure that they function as intended before deployment. Given the multi-domain nature of CPS, it is more appropriate to use a heterogeneous simulation environment to study system dynamics. In this paper, we design and implement an integrated modeling and co-simulation toolchain, called HybridSim, for the design and simulation of CPS. Firstly, HybridSim can transform and import existing system components from multi-domains into SysML, which enables systems engineers to design CPS with only these imported SysML blocks. Secondly, HybridSim can generate Functional Mock-up Units (FMUs) and configuration scripts directly from SysML designs. Finally, HybridSim can co-simulate these FMUs according to the Functional Mock-up Interface standard to synchronize their corresponding simulators and exchange information between them. We demonstrate the convenience and efficiency of HybridSim using a comprehensive hydronic heating system model for Smart Buildings as the case study to investigate the impact of packet loss and sampling rate introduced by the communication network.