PERFORMANCE EVALUATION OF CONVENTIONAL PID CONTROL TUNING TECHNIQUES FOR A FIRST ORDER PLUS DEAD TIME BLENDING PROCESS

Dead time is common to real-time processes and occurs when the process variable does not acknowledge any changes in the set-point. This delay may be due to extensive transportation, imprecise instrument calibration and complex non-linearities present at Final Control Element. First Order Plus Dead Time models make the simulations of these processes, tuning of controllers easier and aid in obtaining the most optimum response. Two orders of transfer functions (fourth and seventh) representing blending systems are modelled as First Order Plus Dead Time using the two points method of approximation. A conventional PID controller is used for both the models. In this work, PID control tuning techniques such as Integral of Weighted Time Absolute Error, Internal Model Control, Ziegler-Nichols and Cohen-Coon, are analysed for the optimum design with the aid of time domain analysis. The responses for all tuning methods are simulated using Simulink in Matlab software. The results indicated that Internal Model Control is the best tuning technique in terms of quick settling, minimum overshoots at the initial stages of the response, minimum rise time and minimum amplitude at the peak time, thereby providing most accurate and robust responses for both the orders of transfer functions.