Enhancing Transient Response and Stability in Automatic Voltage Regulator Systems through Artificial Neural Network-Based Controllers

The study analyzes the use of Artificial Neural Networks (ANNs) in automatic voltage regulator (AVR) systems, focusing on their better control and tuning flexibility over PID controllers. This paper explores employing ANNs in AVR systems, which offer more control flexibility than PID controllers. ANN’s input, activation function, and output parameters permit more accurate control but complicate customization. An AI-based optimization technique finds the best ANN improvements to improve transient response and system stability. This ANN-based methodology is compared to other AI methods’ dynamic performance to prove its usefulness. Pole/zero and Bode analyses assess stability. Finally, the optimized AVR system’s resistance to significant component time constant fluctuations shows enhanced performance and stability. Based on the ANN methodology, the proposed method achieves 59.82%, 56.09%, 14.94%, 34.24%, 35.70%, 21.64%, 12.0%, 41.33%, 14.84% and 15.17% less overshoot than differential evolution (DE).