Diagnostics of machining center by means of neural networks

Abstract

This paper deals with a usage of neural networks for diagnostics of machining center. The diagnostics of machines is very important. To discover backlash, clearance or allowance during the machine work can prevent the mass production of rejects. The aim is to propose a less expensive solution than currently sophisticated commercial software can offer. This paper will explain the diagnostics on specific machining center and its mechanical systems, currently from the point of view - whether there is or there is no error. The introduction into machining center diagnostics is followed by results carried out during simulations with usage of feed forward neural net.