A data-driven method for the development of system frequency response models for frequency stability analysis

Modern power systems are characterized by reduced inertia and primary frequency response as a result of the replacement of conventional synchronous generators (SG) with converter-interfaced renewable energy sources, deteriorating frequency stability. In this context, a novel data-driven methodology is proposed to derive an equivalent aggregated system frequency response (SFR) model that is capable of simulating the power system frequency response following a disturbance. The methodology utilizes active power and frequency response measurements to derive the SFR model through a nonlinear least squares optimization approach. The accuracy of the proposed method is validated by Monte Carlo simulations conducted on the IEEE 9-Bus test system, under both transient events and normal operating conditions. The validation is based on two main aspects. Initially, the model parameters estimated using the proposed data-driven approach are compared with those obtained through analytical calculations. Further, the effectiveness of the proposed approach is evaluated by determining the frequency response of the examined power system under varying types and amplitudes of disturbances. Results verify that in all scenarios the proposed approach provides results similar to those obtained via detailed non-linear dynamic simulations.