Eventos Anais de eventos
COBEM 2017
24th ABCM International Congress of Mechanical Engineering
Transmission Sizing of a Formula-Type Vehicle Using Artificial Neural Networks
Submission Author:
Bruno Diniz , BA , Brazil
Co-Authors:
Leonardo Lopes Nascimento, Bruno Diniz
Presenter: Leonardo Lopes Nascimento
doi://10.26678/ABCM.COBEM2017.COB17-0269
Abstract
The function of a transmission is to optimize the torque of the engine to the wheels, but today there is no defined methodology that analyzes all the parameters and how they relate. For high performance cars, the best condition for transmission is that it can carry the most torque optimally from the engine to the wheels, because on this way the car has greater wheel traction and greater acceleration, but this directly influences increasing fuel consumption, reducing the final speed of the car, and modifying several other parameters of the car. Knowing that the performance ratio is inversely proportional to the consumption and other factors, it is necessary to create an algorithm based on Artificial Neural Networks (ANN) to size a transmission from the priorities and the objectives of the user. In order to perform this work, the data were collected from a vehicle of the SAE formula type because it is a high performance competitive car, where the contour conditions need to be more critical and the parts must be designed with a lower safety factor. A function was developed from the ANN training of MLP architecture, and from there, the semi-axes, the pinion, and the transmission system crown were designed.
Keywords
transmission, Artificial neural networks, vehicle
