Identificação e Controlo de um Conversor de Energia das Ondas usando Modelos Fuzzy e Redes Neuronais

João José Alves Casanova

Key information

Authors:

João José Alves Casanova (João José Alves Casanova Fernandes)

Supervisors:

Duarte Pedro Mata de Oliveira Valério (Duarte Pedro Mata de Oliveira Valério)

Published in

11/15/2011

Abstract

This work is based on the, simplified, mathematical model, of a device designed for extracting energy from waves, which was developed by the Wave Energy Centre, and its goal is the modeling of this device through the models obtained by two alternative technologies, Neural Networks and Fuzzy Logic, with the intended purpose, simultaneously, of making the comparison between the respective levels of performance. The more detailed knowledge of the characteristics of the device, as well as its performance, allows to perform specific control operations in order to optimize its performance, reflected in the optimization of the quantity and quality of the extraction of wave energy. A major premise of this work is to obtain direct and inverse models, with the best quality possible, to enable the design of a controller of the IMC (Internal Model Control) type. The intense and hard research, with a strong computational component, developed during this project was not fully achieved, however, since an inverse model was not successfully achieved. However, the quality and robustness evidenced by direct models, especially by fuzzy models, makes it possibility to opt for a control mechanism of the predictive type, a solution that would be interesting to explore.