PID controller tuning performance evaluation based on integral time square error for coupled tank system

Abstract

This study presents a conclusive performance evaluation based on integral time between the conventional and optimization based proportional integral derivatives (PID) tuning method. The tuning method of interest are including trial and error, auto-tuning, Ziegler-Nichols (ZN), Cohen-Coon (CC), and particle swarm optimization (PSO). The coupled tank system (CTS) is used for the system under consideration as it’s one of the popular technologies in industrial control application. Previous study had compared the tuning performance in terms of its transient response. The transient consists of several parameters such as rise time (Tr), settling time (Ts), peak time (Tp), steady state error, overshoot, and steady state error. Due to that, a conclusive performance comparison could not be achieved. Hence this study proposed integral time squared error evolution which is based on only one parameter which also reflects good overall transient response performance. The results show that of all tuning methods, PSO provide the smallest integral time square error (ITSE) value, while trial and error provide the highest with a value of 12.84 and 203.10, respectively. The ITSE also reflects the transient response performances.