Modelling control the gears with straight teeth by using the neural networks

Show full item record

Files in this item

PDF a_35BBB341.pdf 298.8Kb

Pour citer ce document :
URI: http://hdl.handle.net/2042/52293
Title: Modelling control the gears with straight teeth by using the neural networks
Author: BLOUL, Benattia; BOURDIM, Abdelghafour; BOURDIM, Mokhtar
Abstract: Abstract of this article, the neural networks were applied to determine the control template of the gears size circle involute function angle. Neural networks are appropriate for the diagnosis of complex Alvezaawih systems and control because of its ability to handle data input and output without the need for analytical details of the system. First, we must know the angle to wipe the edge of a tooth surface, with a coordinate measuring machine. We acquire the points entered in standard path of the tooth. Then we execute the neurons on these points to determine the weights is the step that is called learning networks. That last request 94.41s. So it becomes all models of the gear standard with the same characteristics. So, if we wanted a control gear of the same module, it will be only a few points seized by CMMs. On the other hand, is looking for the model of the wheel to control the proper starting weight calculated from previous models. His duration of control esteem 1,6 s. Moreover, we can evaluate the deformities of gear compared to model neurons and the points entered. Finally, the role of this work is the set of model control equipment; we know it contains multiple paths and multivariate parametric complex. So, it's not easy have the control. However, ours, we can know all the deformities of the gears in about 1.60s instead of 45 min at least by the test bench, accordingly, we can say that we achieve gear control the automation.
Subject: CMMs; neural networks; model of controlled gear; defect shape; Conception robuste (S9)
Publisher: AFM, Maison de la Mécanique, 39/41 rue Louis Blanc, 92400 Courbevoie, France(FR)
Date: 2013-12-20

This item appears in the following Collection(s)

Show full item record





Advanced Search