Tampere University of Technology

TUTCRIS Research Portal

Experimental analysis on the turning of aluminum alloy 7075 based on Taguchi method and artificial neural network

Research output: Contribution to journalArticleScientificpeer-review


Original languageEnglish
Pages (from-to)429-437
Number of pages9
JournalJournal Europeen des Systemes Automatises
Issue number5
Publication statusPublished - 2019
Publication typeA1 Journal article-refereed


This paper mainly aims to disclose the effects of cutting conditions on the turning of aluminum alloy 7075 (AA7075). First, the artificial neural network (ANN) was programmed to investigate how cutting parameters, namely cutting speed, feed rate and depth of cut, affect the surface roughness of AA7075. Then, the taguchi method was introduced to design an L27 orthogonal array, in which each cutting parameter is considered on three levels. The results of orthogonal analysis were used to train the ANN called backpropagation neural network (BPNN) on MATLAB. The trained network was applied to predict the surface roughness of AA7075 through MATLAB simulation. Meanwhile, an experiment was conducted under the same conditions. The experimental results were found consistent with the simulation data, indicating that the BPNN is suitable for simulation the turning of AA7075. It is also learned that the cutting speed has the greatest impact on surface roughness; the surface roughness is negeatively correlated with feed rate; the negative correlation is positively mediated by the cutting speed.


  • Artificial neural network (ANN), Cutting speed, Depth of cut, Feed rate, Machining, Surface roughness, Taguchi method, Turning

Publication forum classification

Field of science, Statistics Finland