Towards Comprehensive Modeling of CPSs to Discover and Study Interdependencies.

To a large extent, modeling Cyber-Physical systems (CPSs) and interdependency analysis collaborate in the security enhancement of CPSs and form the basis of various research domains such as risk propagation, attack path analysis, reliability analysis, robustness evaluation, and fault identification. Interdependency analysis as well as modeling of interdependent systems such as CPSs rely on the understanding of system dynamics and flows. Despite the major efforts, previously developed methods could not provide the required knowledge as they have either followed data-driven or physics-based modeling approaches. To fill this gap, we propose a new modeling approach called BG2 based on Graph theory and Bond graph. Our proposed method is able to portray the physical process of CPSs from different domains and capture both information and commodity flows. Based on the fundamental characteristics of the Graph theory and Bond graph in the BG2 model, we discover higher order of dependencies in CPSs and analyze causal relationships within the system components. We illustrate the workings of the proposed method by applying it to a realistic case study of a CPS in the energy domain. The results provide valuable insight into the dependencies among the system components and substantiate the applicability of the proposed method in modeling and analyzing interdependent systems.