Dahbi, S. and Ezzine, L. and Moussami, H.E. (2016) Modeling of surface roughness in turning process by using Artificial Neural Networks. In: UNSPECIFIED.

Full text not available from this repository.
Official URL: https://www.scopus.com/inward/record.uri?eid=2-s2....

Abstract

Surface roughness quality influences significantly the machine parts during their useful life. In fact, a low surface roughness improves the tribological properties, fatigue strength, corrosion resistance, and esthetic appeal of the product. In this paper, we present the modeling of average surface roughness Ra in turning of AISI 1042 Steel at four turning parameters: cutting speed, feed rate, depth of cut, and tool nose radius. The experimental data were collected by conducting turning experiments on a Computer Numerically controlled CNC lathe. By exploiting these data and the Artificial Neural Networks ANNs approach, we developed a Multi-Layers Perceptron MLP with the architecture (4-5-1) and trained with the Levenberg-Marquardt back-propagation algorithm. This network achieved our objectives for the three performance criteria: Correlation Coefficient R2>95, Mean Squared Error MSE<0.1 and average absolute error e<10. Therefore, ANNs are a powerful tool for modeling the surface roughness in turning process. © 2016 IEEE.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Uncontrolled Keywords: Backpropagation; Backpropagation algorithms; Corrosion resistance; Errors; Grinding (comminution); Mean square error; Neural networks; Turning, Average absolute error; Average surface roughness; Correlation coefficient; Levenberg Marquardt back propagation algorithms; Mean squared error; Modeling of surface roughness; Performance criterion; Tribological properties, Surface roughness
Subjects: Business, Management and Accounting
Divisions: SCIENTIFIC PRODUCTION > Business, Management and Accounting
Depositing User: Administrateur Eprints Administrateur Eprints
Last Modified: 31 Jan 2020 15:44
URI: http://eprints.umi.ac.ma/id/eprint/1678

Actions (login required)

View Item View Item