A Comparative Study of PI/PID Classical and Intelligent Tuning Methods

Abstract

Proportional-plus-Integral-plus-Derivative (PID) controllers are widely used in the industrial world. Due to their popularity, a large number of methods have been suggested in the literature to tune them. This paper summarizes and compares seven different PID tuning techniques (in addition to some of their variations). The comparison includes some of the earlyapproaches; namely, Ziegler-Nichols methods, minimum error, phase and gain margins, pole allocation, and LQR. Modern techniques (fuzzy logic and genetic algorithm) are also used and assessed. The seven tuning alternatives are evaluated in terms of their performance, complexity and flexibility.