An algorithmic approach based PID tuning of a real time process

Abstract

A stand-alone control algorithm used to tune the linear process in a classical way is usually called a Proportional Integral Derivative (PID) controller. PID controller, being the most widely used conventional controllers in industrial applications, needs efficient methods to control the different parameters of the process plant. This thesis emphasizes that the conventional approach of PID tuning is not very efficient due to the presence of non-linearity in the real-time process. Non-linearity in the real-time process decreases the stability of the system and makes the process unstable. The output performance of the conventional PID controller has somewhat high peak overshoot and steady state settling time. The main focus of this research is to apply the specific soft-computing and optimization techniques to design and tuning of the PID controller to get an output with better static and dynamic performance. The controller design for open-loop unstable systems is more difficult than that of the stable processes and the difficulty increases further when there exists a time delay. The performance specifications that are usually achieved for stable systems are difficult to be achieved for an unstable system. Because, for unstable systems, the performance specifications such as overshoot, settling time are larger and there exists a minimum and maximum value of controller gain below which and above which the closed loop system cannot be stabilized. These two values of maximum and minimum controller gain narrow down as the time delay increases restricting the performance of the closed-loop system. The dynamics of many processes can be represented by first or second order processes plus time delay. There are wide collections of control techniques that can be useful to encounter the control objective of the system and these depend on the factors of which the suggested design objective might depend on. newline