Compartir
Título
Deep Convolutional Neural Network Modeling and Laplace Transformation Algorithm for the Analysis of Surface Quality of Friction Stir Welded Joints
Autor(es)
Materia
Machine Learning
Friction Stir Welding
Convolutional Neural Network
Surface Defects
Laplace Algorithm
Fecha de publicación
2021-10-05
Editor
Ediciones Universidad de Salamanca (España)
Citación
ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal, 10 (2021)
Resumen
The quality of Friction Stir Welded joint depends on the input parameters like tool rotational speed, tool traverse speed (mm/min), tool tilt angle, and an axial plunge force. If there is any variation in these input parameters then there will be a chance of formation of various surface defects such as groovy edges, flash formation, and non-homogeneous mixing of alloys. The main objective of the present work is to use machine learning algorithms such as Deep Convolutional Neural Network (DCNN) and Laplace transformation algorithm to detect these surface defects present on the Friction Stir Welded joint. The results showed that the used algorithms can easily detect such surface defects with good accuracy.
URI
ISSN
2255-2863
Colecciones