Comparative study between ZN PID, GA based PID and ANFIS PD controller to a shell and tube heat exchanger temperature control system

Abstract

A heat exchanger device is widely used in process industries because it is capable of sustaining a wide range of temperatures. The main purpose of a heat exchanger is to transfer heat from a hot fluid to a cooler fluid so that the temperature of the process fluid is controlled. Among the different heat exchanger types, shell and tube heat exchanger is the most commonly used type in the process industry. The heat exchanger temperature control system is a highly nonlinear, uncertain, time-delayed, and complex system. In addition to that, it is accompanied by the presence of two predominant disturbances, namely flow variation and temperature variation of input fluid. In this thesis, the performance of Ziegler–Nichols based proportional integral and derivative controller, genetic algorithm based proportional integral and derivative controller, and adaptive neuro fuzzy inference system has been analyzed and compared under four scenarios, namely no disturbance, flow variation disturbance, temperature variation disturbance, and both disturbance conditions. Among the proposed controllers adaptive neuro fuzzy inference system has provided the best transient response performance with and without the presence of disturbance effects. The settling time of the adaptive neuro fuzzy inference system for the proposed system under no disturbance, flow disturbance, temperature disturbance, and both disturbance conditions is 25.045sec, 25.439sec, 26.948sec, and 27.932sec. It has been improved by 54.9%, 55.2%, 52.8%, and 52.1% when compared to the Ziegler–Nichols tuned proportional integral and derivative controller. And the percentage overshoot of the adaptive neuro fuzzy inference system for the proposed system under no disturbance, flow disturbance, temperature disturbance, and both disturbance conditions is 0.901%, 0.462%, 0.1226%, and 0.00012% %. On the other hand, among the three objective functions used the overall performance of genetic algorithm based proportional integral and derivative controller tuned based on minimizing time integral of absolute error objective function is better than the one tuned based on minimizing integral of absolute error and integral of squared error objective functions.