Real-Time Damping of Low-Frequency Oscillations in Power System Networks Using Machine Learning

Stability is a fundamental requirement for electrical power systems to function safely, efficiently, and continuously. Common in interconnected power networks, low-frequency oscillations (LFOs) present serious stability issues and necessitate sophisticated mitigating techniques. In order to improve system reliability and minimize undesired oscillations, the power system stabilizer (PSS) must have its parameters optimized. This study offers a fresh approach to doing so. The first is a single-machine infinite bus power system, and the second is a unified power flow controller (UPFC) device. The proposed ensemble model is based on the Harris Hawks Optimization algorithm (HHOA) and is formed using the extreme learning machine (ELM), a machine learning (ML) technique. The analysis of eigenvalue and minimum damping ratio analyses under different loading conditions shows how successful the proposed approach is. Additionally, the time-domain simulation comparison shows how much better the proposed approach is than the conventional one. Furthermore, the recommended method's ability to foresee power system stabilizer parameters and so alleviate low-frequency oscillations in real-time is given credence by the good values of the statistical performance measures.