Ajuste automático de controladores PID utilizando técnicas computacionais inteligentes

Abstract

This work presents a comparative study of results obtained from the application of intelligent systems in an automatic tuning algorithm of PID controller parameters. The plants used in the simulations include generic dynamical systems of third, fourth and fifth orders and an alcoholic fermentation process. For applying in this algorithm, the dynamic systems have to be stable for proportional gain near zero and achieve sustained oscillation with its increase. In this context, setting routines were developed using Fuzzy System and Fuzzy Cognitive Maps as a tool for decision making in adapting parameters process. It was presented the foundation on the PID control and tuning techniques, dynamic systems and intelligent systems. After the development of the routines, the computer simulations were applied using Matlab® Simulink, using all dynamic systems for each intelligent tool. Finally, the results of the simulations are presented, which proved the ability of the adaptive algorithm by comparing the efficiency of the application of both intelligent systems.