Document Type : Original Article
Authors
1
School of Mechanical Engineering,, Iran University of science and Technology
2
School of Mechanical Engineering, Iran University of Science & Technology, Narmak, Tehran, Iran
3
Assistant Professor of Mechanical Engineering School of Mechanical Engineering Iran University of Science and Technology (IUST)
10.22034/ijme.2024.412021.1816
Abstract
Considering the importance of surface roughness in machined parts,، it is necessary to identify the factors that play an essential role in determining and improving surface quality. In this regard, they can be predicted and optimized by choosing suitable parameters. One of the methods used in this area is the prediction of surface roughness with traditional methods. However, limited studies were found on using meta-heuristic methods. As a result, this study proposed a model using A.I. algorithms and, more specifically, the fuzzy neural network algorithm and the combined algorithm of the fuzzy neural network. In total, 162 tests were used on three aluminum alloys 7075, 6061 and 2024, which include a variation of cutting speed, depth of cut, tool coating, feed rate, and surface roughness output. The fifth input parameter was the mechanical properties of materials, such as tensile strength, shear strength and hardness. During the modeling process, the corresponding data of two alloys were used as training data and the third as test data. The simulation performed on the data was evaluated using regression (R) and root mean square error (RMSE) criteria. The most accurate result among the two ANFIS and ANFIS-GA methods was obtained with a regression of 0.838 for the surface roughness of the test data.
Keywords
Main Subjects
)