Effect Dead Time On The Design Controller In Industrial Processes

The application of deadtime compensation techniques is often limited by the need for an accurate model of the process dynamics and gain. This paper reviews three deadtime compensation techniques which are available in industrial control systems for the control of deadtime dominant processes; Predictive PI controller (PPI), PIDd, and Variable Horizon Predictor. The basic approach taken by these techniques in providing deadtime compensation is compared to the Smith predictor and the Dahlin algorithm. It is shown that the response of the Dahlin, PPI, and PID d algorithms may be achieved using the Smith Predictor. The sensitivity of these techniques to process and model mismatch is explored. The approach used by the Variable Horizon Predictor in reducing sensitivity to process model mismatch is detailed. The Variable Horizon Predictor control is shown to be more robust than PID for changes in process dynamics.

Keywords:

PID Controller , Predictive PI Controller , Deadtime