An approach for load frequency control enhancement in two-area hydro-wind power systems using LSTM + GA-PID...

Area of Science:

• Electrical Engineering
• Control Systems
• Renewable Energy Integration

Background:

• Interconnected power systems face challenges with load disturbances, impacting frequency stability.
• Traditional PID controllers struggle with nonlinearities and uncertainties in hydro-wind power systems.
• Existing control strategies often result in slow response times and significant overshoot.

Purpose of the Study:

• To develop an advanced Load Frequency Control (LFC) strategy for two-area hydro-wind power systems.
• To improve control precision and dynamic adaptability in the face of load disturbances.
• To overcome the limitations of traditional PID controllers in complex power system environments.

Main Methods:

• Implementation of a hybrid Long Short-Term Memory (LSTM) neural network and Genetic Algorithm-optimized PID (GA-PID) controller.
• Utilizing LSTM for disturbance forecasting based on historical data and gradient descent.
• Employing GA for real-time optimization of PID controller parameters.
• Performance evaluation through MATLAB/Simulink simulations and hardware validation.

Main Results:

• The LSTM+GA-PID controller demonstrated a 2.33-fold reduction in settling time versus GA-PID and a 4.07-fold reduction versus classical PID.
• Achieved a 3.27% reduction in overshoot and mitigated mechanical power output perturbations by 3.43%.
• Hardware validation confirmed the model's robustness and effectiveness.

Conclusions:

• The proposed LSTM+GA-PID controller offers superior performance for LFC in hydro-wind power systems compared to traditional methods.
• The hybrid approach effectively handles system nonlinearities and uncertainties, enhancing stability.
• This advanced strategy provides a robust solution for maintaining power system frequency under dynamic load conditions.