Automatic Voltage Regulation using Control Systems and LSTM Model

Abstract

In power generation systems, maintaining operational voltage within predefined standards is crucial, as deviations can lead to equipment malfunction and compromise the stability of the power supply. Traditional voltage regulation, often managed manually by onsite personnel, introduces uncertainties and inefficiencies, hindering the economic and stable operation of power facilities. To address these challenges, this paper introduces an advanced, computer-controlled Automatic Voltage Regulator (AVR) system. Utilizing sophisticated control theories such as Proportional-Integral-Derivative (PID) dynamics and adaptive control algorithms, the AVR enhances the precision of voltage adjustments in real-time. This paper innovates further by integrating a Long Short-Term Memory (LSTM) model to predict voltage fluctuations and dynamic system loads more accurately. This predictive capability allows for preemptive adjustments, optimizing the voltage regulation process to ensure stability while considering the economic efficiency of power system operations. Through comprehensive simulations and robust hardware integration, the proposed AVR system not only promises enhanced operational reliability and efficiency but also significantly reduces the need for manual intervention, paving the way for fully automated power system management.