PhD defence of Yao Jie Cai – Locate Power System Forced Oscillations based on Sparse Identification of Nonlinear Dynamics

Abstract

Forced oscillations (FO), resulting from periodic external disturbances, pose significant threats to the security and stability of the power system. The increasing penetration of Inverter-Based Resources (IBRs) introduces additional challenges in identifying and locating these oscillation sources. To address this issue, this thesis proposes a novel, purely data-driven method that is based on sparse identification of nonlinear dynamics (SINDy) for online forced oscillation source location. Unlike previous approaches, a unified representation of forced oscillations originating from IBRs is considered, enhancing the algorithm's effectiveness in modern power systems. The proposed method requires no model information and operates with minimal tuning and low computational cost, making it viable for online applications. Extensive validations on both simulated cases (WECC 179-bus and 240-bus systems) and actual oscillation events (ISO New England system from the IEEE Task Force test cases library) demonstrate the algorithm's capability to accurately identify oscillation sources, even under resonance conditions and poorly damped natural modes. The results confirm the robustness and suitability of the method for real-world power system applications.