Sintonia multiobjetivo de controlador preditivo generalizado por NSGA-II

Abstract

Controll systems are present in routine of most people. For example, in a lift there are a devices which control speed, aceleration, and position; where based on input they must obtain a proper output. The function and performance of such devices are strictly related to the parameters of each kind of controllers. The choosing of these parameters is known as tunning. There are several tunning methods. On one hand, the most common are those where the parameters are defined beforehand and keep fixed allong all the process. On the other hand, there are automatic parameter tunning, called on-line tunning, that are common for complex processes. This work focuses on the use of multi-objective optimization applied to the parameter tunning on a generalized predictive control. The proposed solution is based on the Non-dominated Sorting Genetic Algorithm II – NSGA-II, which is caracterized by dealing with the optimization of multiple objectives simultaneously. From the results we can observe that the NSGA-II algorithm reaches promissing solutions that allow a better decision making.