Model Predictive Control Education using a Rapid Prototyping Industrial Platform.

The recent advances in implementation methods and in the efficiency of computing devices increased the interest of industry in using model predictive control (MPC) methodologies in control applications characterized by fast dynamics. In order to cope with such a growth of interest, control designers who are adequately trained for the MPC design practice are needed. In this regard, suitable experimental activities must be set up inside the educational programmes involving MPC subjects. In this context, rapid prototyping (RP) environments play a role of primary relevance in testing and performing MPC design iterations on a real-time system. This paper introduces a flexible RP tool for MPCs developed on a National Instruments industrial device. It is aimed at allowing students with basic knowledge of predictive control theory, to train their ability in the design of MPC controllers by means of a 'learn-by-doing' approach. Experimental practice on a magnetic lévitation system as well as a student assessment and survey data are presented to show the effectiveness of the proposed MPC RP environment. [ABSTRACT FROM AUTHOR]