Fuzzy logic based high performance PID controller

Abstract

To start with, practicing engineers generally use Ziegler-Nichols tuned PID (ZNPID) controllers in most of the process control loops. However, to achieve satisfactory performance ZNPID is manually fine tuned using operator's knowledge. To avoid such manual involvement, this paper proposes a fuzzy logic based High Performance PID Controller (HP-PID) incorporating operator's strategy in running a plant. In HP-PID, depending upon the process trend the proportional and integral gains of ZNPID are continuously updated by an online gain modifier `α', which is derived by firing 25 fuzzy rules defined on the process error and change of error. Performance of the proposed HP-PID under both set point change and load disturbance is observed for a number of linear and nonlinear higher order processes with considerable changes in process dead time. It shows noticeably improved performance compared to ZNPID and augmented ZNPID (AZNPID) recently reported in the leading literature.