A Two-Level Distributed Approach to Power Network Modeling

Power network modeling is traditionally performed in a centralized way, where the network components and topological relationships are manually modeled at control centers. However, in large-scale power systems, this conventional modeling approach can suffer from poor efficiency and inevitable human errors, which may affect the reliability of system models. Currently, the development of distributed applications requires detailed modeling of local systems and consistency between local and global models, which is not easy with conventional modeling. This paper presents a two-level distributed modeling approach that reflects the demands of distributed applications. It describes the processes of substation modeling, model transformation, and control center modeling. Using this modeling approach, substation models can be derived in a distributive manner at substations, while system models are automatically generated at control centers. This novel modeling approach not only adapts to the future architecture of power system operation, but it also overcomes the drawbacks of conventional modeling. In addition, it has been applied to a real power system, where it demonstrated high-efficiency performance.