Dynamic Clustering of Operating Points for Online Equivalent Modeling of Interconnected Power Grids with Renewable Energy

As renewable energy sources become increasingly integrated into interconnected power networks, system operating points (OPs) undergo frequent and unpredictable shifts. However, conventional delays in updating equivalent model parameters during these OP transitions often compromise modeling accuracy. To address this challenge, this study proposes an online dynamic OP clustering method for interconnected grids featuring wind and photovoltaic generation. First, an equivalent model for renewable-integrated interconnected grids is established. Subsequently, a dynamic OP clustering strategy is developed; this strategy combines an offline construction phase utilizing joint probability distributions and data clustering with an online update mechanism that dynamically adjusts cluster boundaries via membership calculations. This approach enables real-time clustering, effectively minimizing equivalence errors and adapting swiftly to ongoing network variations. Simulation results based on the China–Mongolia interconnected power grid demonstrate that the proposed method significantly outperforms traditional static approaches in both equivalence accuracy and computational adaptability. By delivering precise, real-time network equivalents, this approach provides robust support for practical grid operations, including online security assessment, optimal power dispatching, and transient stability analysis, thereby contributing to the long-term stability and sustainability of modern power systems.