Usinagem  - USI 63

Title  :

INFLUENCE OF THE FORCE COMPONENTS ON THE PERFORMANCE OF ARTIFCIAL NEURAL NETWORKS FOR TOOL WEAR MONITORING

Abstract :
A large number of works have been presented describing the results from research with Artificial Neural Networks (ANNs) for tool wear monitoring. In these works the ANNS were “fed” with different types of information, frequently with values from measurements of force and acceleration. The correlation between these values and the tool wear state have been shown in the literature but their influence in the performance of the ANNs, and most of all, the correlation between the results from the ANNs and the tool wear were not clearly presented. In the present work the authors show the results of experiments and analysis to find out the force components that gives more contribution for the performance of the ANNs tested and trained with data from tool wear tests. These data were sampled during the turning of non-ductile steel bars, with different types of tools and several cutting conditions.
Autores :
Canabrava Filho, José S.
Barrow, George
Trabalho Completo :


    1o. Cobef | Comissão Organizadora | Palestras | Sessões Técnicas | Autor | Revisores