Real-time regulation of the terminal voltage in synchronous generators requires a high-performance Automatic Voltage Regulator (AVR) system. Enhancing its dynamic performance remains an ongoing challenge, with conventional Proportional–Integral–Derivative (PID) and Fractional–Order PID (FOPID) controllers widely used due to their simplicity and effectiveness. However, their limited parameter space often constrains performance improvements. This study presents a data-driven control structure that integrates a variation of neural network, the Cerebellar Model Articulation Controller (CMAC), in a parallel feedforward path alongside PID and FOPID controllers in the feedback loop. The objective is to improve transient response and tracking accuracy by minimizing settling time and overshoot. To optimize the increased number of control parameters introduced by the CMAC network, a Modified Safe Experimentation Dynamics Algorithm (MSEDA) is employed. This single–agent heuristic method is suitable for data-driven control frameworks with high-dimensional tuning parameters, as it exhibits low computational complexity. The proposed structure is evaluated through the convergence of the Figure of Demerit (FOD), time-domain responses, and robustness under measurement noise, external disturbances, and parameter variations. Simulation results confirm that the CMAC–based PID and FOPID controllers significantly outperform benchmark PID, FOPID, and PIDA controllers, yielding lower values across all performance indices.
Saat et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: