FA-Based Optimal PIDA Controller Design for AVR System

Abstract

Firefly algorithm (FA) was firstly proposed by Yang in 2008 as one of the most powerful population-based metaheuristic optimization techniques for solving the hardest optimization problems including real-world engineering practices. In this paper, the FA is applied to design an optimal proportional-integral-derivative-accelerated (PIDA) controller for the automatic voltage regulator (AVR) system. Based on the optimization context, the sum-squared errors between referent input voltage and actual terminal output voltage of the controlled system are performed as the objective function to be minimized. As results, the optimum PIDA controller for the AVR system is successfully obtained by the FA. Moreover, the FA-based design approach performs high robustness once parameter variations are occurred.