Modeling and control of automatic voltage regulation for a hydropower plant using advanced model predictive control

Fluctuating voltage levels in power grids necessitate automatic voltage regulators (AVRs) to ensure stability. This study examined the modeling and control of AVR in hydroelectric power plants using model predictive control (MPC), which utilizes an extensive mathematical model of the voltage regulation system to optimize the control actions over a defined prediction horizon. This predictive feature enables MPC to minimize voltage deviations while accounting for operational constraints, thereby improving stability and performance under dynamic conditions. The findings were compared with those derived from an optimal proportional integral derivative (PID) controller designed using the artificial bee colony (ABC) algorithm. Although the ABC-PID method adjusts the PID parameters based on historical data, it may be difficult to adapt to real-time changes in system dynamics under constraints. Comprehensive simulations assessed both frameworks, emphasizing performance metrics such as disturbance rejection, response to load changes, and resilience to uncertainties. The results show that both MPC and ABC-PID methods effectively achieved accurate voltage regulation; however, MPC excelled in controlling overshoot and settling time—recording 0.0 % and 0.25 s, respectively. This demonstrates greater robustness compared to conventional control methods that optimize PID parameters based on performance criteria derived from actual system behavior, which exhibited settling times and overshoots exceeding 0.41 s and 5.0 %, respectively. The controllers were implemented using MATLAB/Simulink software, indicating a significant advancement for power plant engineers pursuing state-of-the-art automatic voltage regulations.