Performance Analysis of the Reduced Order Models of Automatic Voltage Regulator (AVR) System

The Automatic Voltage Regulator (AVR) plays a unique role in the functioning of the electrical grid. The AVR's complexity stems from the fact that its transfer function is of a higher order. Therefore, the analysis of higher order AVR systems results in a challenging task. Hence, the lower order models of AVR have been developed and examined in this work. Pade's approximation, Big Bang Bing Crunch, and Particle Swarm Optimization are utilised to acquire reduced order models of the AVR system. In addition, the simplified AVR models have had a PID controller developed for them utilising both IMC and the conventional Tyreus-Lyuben (TL) techniques. Time response characteristics, stability margins, and performance indices have been evaluated between the TL-PID and IMC-PID controllers. Results from a comparison of the proposed PID controllers show that lower-order models of the AVR system function well in comparison to their higher-order counterparts.