Data-Driven Inertia Evaluation of Power Systems Based on Electromechanical Response Characteristics

Inertia evaluation is a fundamental prerequisite for maintaining the stable operation of power systems with high renewable penetration. Accurate inertia estimation not only helps determine the safety status of the system but also provides guidance for optimal allocation of inertia support resources. This paper proposes a data-driven method for regional effective inertia evaluation by analyzing the coupling relationship between electromechanical response characteristics in the frequency domain and system inertia, and deriving an inertia evaluation model accordingly. A stochastic subspace identification algorithm is employed to extract electromechanical features and calculate system inertia, while a threshold regulation strategy is adopted to avoid numerical singularity problems. The proposed method demonstrates good identification accuracy and practicability. Case studies using Digsilent-generated time-domain simulation data for a single-machine infinite bus system validate the effectiveness of the proposed method.