Meta-Heuristic Approach to Design Controller for AVR System

In this paper, Firefly Algorithm (FA) based design of the optimal PID controller is proposed for the benchmark Automatic Voltage Regulator (AVR) system. The best possible solution (Kp, Ki, Kd) for the PID controller is obtained by minimizing the multi-objective cost function. In this work, the FA explores the three dimensional search space, till it discovers the optimal PID parameters. This paper also presents a comprehensive study various search assisted FA, such as Brownian, Levy and chaotic exploration. The simulation work is implemented using the traditional and modified form of PID controllers. The performance of PID is computed based on the time domain values and the error values. The performances of chosen FA are also assessed based on its run time. From this research work, it can be observed that, all the search techniques tender similar time domain and error values. The iteration time taken for the Chaotic FA is relatively lesser than other FAs. Proposed approach is also authenticated against the other related works existing in the literature and it is noted that, FA based PID controller offers better result. Citation: Aishwarya. C. R, St. Joseph's College of Engineering; Andal Santhini. R ,St. Joseph's College of Engineering; M. Durga Devi ,St. Joseph's College of Engineering; Avashya.D. Roy ,St. Joseph's College of Engineering. "Meta-Heuristic Approach to Design Controller for AVR System." Global Research and Development Journal For Engineering : 11 - 16.