Eventos Anais de eventos
DINAME2019
DINAME2019
Trajectory tracking control of a seesaw-propeller system using a feedback-feedforward approach and artificial neural network
Submission Author:
Gabriel de Albuquerque Barbosa Baumann , RN
Co-Authors:
Gabriel de Albuquerque Barbosa Baumann, Gabriel da Silva Lima, Wallace Moreira Bessa
Presenter: Gabriel de Albuquerque Barbosa Baumann
doi://10.26678/ABCM.DINAME2019.DIN2019-0168
Abstract
A controller for a nonlinear dynamic system is proposed in this work. This system is composed of a propeller driven seesaw, whose angular position must be controlled. For this purpose, the ordinary differential equation that describes the system dynamics is shown and the control strategy was developed through a feedback-feedforward configuration combined with neural networks. The approach consists in the Radial Bases Function artificial neural network application in a feedforward compensator, to estimate and predict plant disturbances, and to overcome the feedback linearization technique limitations. Results obtained by numerical simulation of the controller with and without the performance of the feedforward with the neural network compensator are presented and discussed to demonstrate the effectiveness of the approach presented in the project in promoting an approximate state of robustness.
Keywords
control of nonlinear system, feedback linearization, feedforward compensator, artificial neural network
