Design of Intelligent controller for Isothermal Continuous Stirred Tank Reactor System

Abstract

Process control is one of the research areas, for which the design of an ideal controller is required. There are numerous models available in the real time where controllers plays significant and important role. The PID controller works for most control systems in the real time applications. PID controller is an extensively utilized feedback system. It has been said that there are over 95% PIDs at the interval of commercial projects of the specialized department. In this dissertation, soft computing methods have been compared with modifying the PID controller. CSTR is one of the common used reactors in chemical process. These controllers are tuned with different control stratergies to control the concentration of an isothermal CSTR, which is a second-order system with zero right half plane and it additionally understands the concentration of an isothermal CSTR that is prepared with state space condition. The primary goal of the controller is to control the concentration of an isothermal CSTR product, while the flow may be varied. In this dissertation the effect of various controllers like conventional PID controller, Fuzzy Logic Controller (FLC) and Genetic Algorithm (GA) has been presented. The PID controller based on genetic algorithm technique that has been used for different population sizes for various kind of performance indices. The gain value and transient responses for various type of performance indices received from the PID controller by using genetic algorithms provide better results than the conventional PID Controller by using Zeigler-Nichols technique and the fuzzy based controller.