Continuous Human Learning Optimizer based PID Controller Design of an Automatic Voltage Regulator System

Abstract

In this paper, an intelligent design and tuning method for the proportional-integral-derivate (PID) controller of an automatic voltage regulator (AVR) system using a novel heuristic algorithm termed as the continuous human learning optimizer (CHLO) is presented. The CHLO is inspired by human learning mechanisms wherein a well-defined rule based probabilistic procedure of random learning, individual learning, and social learning leads the search process. The CHLO is implemented in Matlab and identification of the PID parameters for said AVR system is done by stating the design task as an optimization problem. The problem statement is formulated to minimize the integral of absolute error (IAE) criterion with in-built weighted preferences for transient response characteristics. The simulation experiments are devoted both towards the application as well as in exploring the behavioral parameters of the CHLO optimizer. The performance measures such as transient respone indices, root locus analysis, and bode analysis are carried out. The obtained results are compared with other heuristic approaches in terms of percentage improvement in transient response indices. The numerical simulation results endorse competitiveness and better optimization potential of the proposed method than the biography based optimization (BBO), particle swarm optimization (PSO), differential evolution algorithm (DEA), and artificial bee colony (ABC) algorithm in parameter identification of the AVR system.