KNOWLEDGE-BASED PID CONTROL

Abstract

The Proportional-Integral-Derivative (PID) controllers are widely used in the process industry and it is foreseeable that it will continue to be so owing to their simplicity. Hence, any new results will certainly bring about much benefits. In this thesis, issues related to the performance and robustness of PID control systems are raised. Specifically, the relative performance between the PI and PID controllers is accessed. The results of the studies will aid in the choices between the simpler PI controllers or the more sophisticated PID controller. This thesis also presents a robustness analysis of well-known PID design methods. For online monitoring of the robustness stability of closed-loop systems, the difficulty of providing a plot of the gain and phase margins is that they cannot be determined easily. In this thesis, a set of non-linear equations are solved analytically such that the gain and phase margins can be calculated on-line and in real-time.