LSTM-based PSS Design for Modern Power Systems

With the ever-increasing incorporation of wind and solar power in electric system, enhanced performance classical bulk hydropower plants for robust operation system is required. This current energy transition may cause a rapid increase undesirable low-frequency oscillations (LFOs) modern operations. A stabilizer (PSS) located at one solution to damp such oscillations. paper presents new method based on Long Short-Term Memory (LSTM) neural networks sine-wave phase shifting possibly enhance PSS damping. The proposed controller considers input rotor speed deviation as damped sinusoidal signal, simplifying control real-time optimization PSSs parameters. Results show that LSTM architecture able learn multiple sine waves with different frequencies decay rates. Therefore, can operate entire range LFOs, unlike simple feedforward network (FNN) controllers, which only function single LFO frequency.