Optimization of Type-2 Fuzzy Controllers
Process control, Fuzzy logic, Type-2 fuzzy controllers, Optimization
Several industrial strategies and control algorithms are already used and presented in the literature. Among the existing techniques, the fuzzy controllers stand out for their ability to deal with nonlinear servers present in real plants and for being able to represent specialist knowledge, which is inaccurate and inaccurate in mathematics. This work studied two types of existing fuzzy controllers, based on the Sugeno model, type-1 fuzzy, here classified as classic fuzzy, and type-2 fuzzy. Due to the difficulty to tune the fuzzy controllers, an alternative is presented in this qualification proposal, which are the optimization algorithms that will be used to refine the tuning of the controllers and, thus, obtain better answers for the studied plants. To validate implementations two plants were used, a servo-DC motor and a system of coupled tanks. In order to compare the behavior of controllers, a PI controller was implemented for each system. In order to quantify and qualify each controller, three evaluation indices are used: the ITEA, IEA and the Goodhart index, the latter used to take into account also the control signal. The results obtained show that the type-2 fuzzy controller obtained better results when compared to the others controllers implemented. This controller presented the lowest indices that the others, although all controllers have an adequate response for both systems.