Eventos Anais de eventos
COBEM 2019
25th International Congress of Mechanical Engineering
USING A NEURAL NETWORK AND GENETIC ALGORITHM FOR A CANTILEVER BEAM WITH SHUNT CONTROL
Submission Author:
Alexandre Victor Brandão dos Santos , MT
Co-Authors:
Alexandre Victor Brandão dos Santos, Venicio Silva Araujo, Guilherme Silva Prado, Heinsten Frederich Leal dos Santos
Presenter: Alexandre Victor Brandão dos Santos
doi://10.26678/ABCM.COBEM2019.COB2019-1958
Abstract
Smart techniques have been receiving widely attention in recent years as an alternative for solve engineering problems.Inside this concept, algorithms of optimization that seeks empirical connections between parameters of a systems stands out, mostly due its nature of dealing with data without constantly resort to complex equations. Therefore, this work aims to the usage of two categories of these algorithms, the Genetic Algorithm and Artificial Neural Network, in intention to establish vibration control in a structural set of a cantilever beam coupled to a piezoelectric layer in the piezo-beam configuration, doing a comparison between them focusing on time’s execution and attenuation provided for the first natural mode of the beam. Were used piezoeletric ceramics of 10, 25 and 50mm of length, setted up since 5mm to 45mm from the crimp. The results shows that Genetic Algorithm and Neural Network presents close answers when they were configured to define electronic parameters for a shunt circuit, presenting average 23dB of damping, the larger difference comes in time of executions, where Genetic Algorithm presents average time of 1233,44s per execution and Neural Network presents 98,11s
Keywords
Vibration control, Genetic Algorithms, Neural Network, Shunt Control, Piezoelectric materials
