A hybrid framework for assessing regional inertia estimation in bulk power systems using COI-driven spectral clustering

Modern power systems are facing growing challenges in frequency stability due to the increasing integration of renewable energy sources. Accurate regional inertia estimation is essential to address this issue. This paper proposes a novel grid partitioning method that enhances spectral clustering to identify regional centres of inertia (RCOIs) using complex network analysis. These RCOIs enable precise inertia estimation through an auto-regressive moving average (ARMAX) model, which combines data-driven and model-based approaches to achieve this. Validated on the NETS-NYPS test system under conventional and renewable generation scenarios and on the NPCC test system under traditional generation, the method achieves estimation errors below 3%, with renewable-specific accuracy nearing 10%. The framework offers a robust solution for monitoring regional inertia, thereby enhancing grid stability, operational efficiency, and resilience in dynamic power systems.