AUTOTUNING OF GAIN-SCHEDULED PH CONTROL - AN EXPERIMENTAL-STUDY

pH control is known to be a difficult control problem. In this work, t he gain-scheduled proportional-integral (PI) controller is used to ove rcome process nonlinearity. Using the gain-scheduled PI controller the nonlinear control problem becomes the identification of the titration curve and the design of the PI controller. With necessary instrumenta tion and appropriate model parameterization, the titration curve can b e found in a robust manner. Autotune identification is employed to est imate the titration curve first and goes on to find the nominal tuning constants. This is the gain-scheduled autotune identification (g-ATV) procedure. The g-ATV procedure is tested in a laboratory scale pH neu tralization system. The results show that robust estimation of the tit ration curve can be achieved from continuous cycling. Furthermore, rel iable and effective nonlinear pH control can be obtained in a realisti c operating condition as shown in this experimental study.