Swarm-Based Optimizers for PID Tuning of Bioreactor

Abstract

This paper addresses the proportional-integral-derivative (PID) controller design for growth stage of bioreactor process. It is second-order, time-delayed, unstable process. The controller design is based on the swarm-based evolutionary algorithms. In this study, particle swarm optimization (PSO), bacterial foraging optimization (BFO), and BFO assisted PSO algorithms are used to find the tuning parameters of PID controller. PSO and BFO algorithms are explained in detail. Effect of acceleration coefficients in PSO is discussed briefly. To validate the effect of acceleration coefficients, PSO algorithm was simulated for different combinations of these coefficients. For better understanding, results of ten individual runs were reported for PSO, BFO, and hybrid PSO + BFO. In this work, comparative study of PSO, BFO, and BFO assisted PSO is presented. The performance comparison includes the error minimization and time complexity analysis.