Resumen:
In this research work, three advanced control algorithms are designed, implemented, and compared to control
level and temperature processes in a virtualized Festo Modular Production System of Automatic Processes (MPS® PA) workstation, with an educational focus. A traditional PID control strategy is designed using the lambda tuning technique for each variable, the second controller designed is a fuzzy PD controller with integral action at the output and finally, a Model Predictive Control (MPC) algorithm is designed. These three controllers have
as objectives, to minimize the error in steady-state and reduce the sudden actions of control, the three control strategies are compared and the maximum overshoot, error in steady-state and settling time are evaluated as performance indicators, moreover, the actuator performance is analyzed. The design of the Festo MPS® PA virtual workstation is also part of this research work. The gamification of the station allows teaching and learning
control techniques under a controlled industrial environment. The designed virtual station has physical and operational characteristics identical to the real station with all its instruments virtualized in 3D using Computer-Aided Design (CAD). The controllers designed for the two variables were implemented in Matlab. The mathematical model of processes is linked to the Unity software through shared memory. The results indicate that
the advanced controllers (diffuse and MPC) obtained the best performance, which allows increasing the actuators’ lifetime.
