On the Influence of the Prediction Horizon in Dynamic Matrix Control

摘要

In this paper authors propose to study the influence of the prediction horizon on Dynamic Matrix Control (DMC). This kind of controllers is a particular type of Model Predictive Control (MPC), which are framed as advanced controllers. To follow a scientific procedure and to objectify such influence, authors have taken into consideration a set of indexes defined to measure the performance reached at the output of a system when it is controlled by different DMC controllers, which have been synthesized varying in a planned and systematic way the value of the prediction horizon. The system on which the influence of the parameter has been analysed has been chosen because it has shown to be difficult to control using classical control schemes, as a discretized Proportional Integral Derivative (PID) control tuned by means of the Ziegler-Nichols method. In this study 720 different experiments have been carried out, and the effect of the prediction horizon parameter is summarized by means a number of figures.