Developing Models And An Understanding Of Their Limitations

Students often do not appreciate the difference between the model of a system and a real system. Typical homework problems often have students apply control strategies on systems (plants) that have been given to them in the problem statement. Utilizing our Education Control Products (ECP) systems, as well as Matlab’s Simulink, we developed a sequence of laboratories to emphasize the difference between the model of a system and a real system in our undergraduate controls classes. In the initial labs the students determine models for the different plants they will be utilizing. These models are created using specialized Simulink blocks we developed that drive the ECP systems. The goal of these early labs is actually to develop a model of the plant to replace the Simulink hardware drivers. Subsequently, controllers and/or observers are designed to meet design specifications utilizing these system models. Then the students replace their model of the plant with the hardware driver (Simulink block) for the ECP system. These Simulink systems, with the student designed controllers/observers, are used to run the real ECP systems. Finally, the students compare the predicted response (utilizing their model) with the actual response of the real system. Our assessments of the students indicate that after the course the students have a better understanding that the response of the model will only approximate that of the real system.