Comparative study of blocking and suppression in predictive control

Abstract

The performance of predictive controllers based on convolution models can be modified in a number of ways to alleviate the problems of “ringing” and excessive controllers actions. At present these modifications are mainly done through the use of the so-called “Tuning Parameters”, penalizing the control actions and suppressing their variations after a specified number of sampling periods. In the present work an alternative method, based on blocking the controller's actions over several sampling periods, is studied and compared to suppression method using simulations with several SISO systems. The proposed blocking method with or without penalizing factor proved to be effective, easy to implement and compared favorably with the other methods.

Keywords

• internal model control
• convolution models
• blocking
• suppression