Affordable Access

Publisher Website

Signal processing in defect detection using back-propagation neural networks

Authors
Journal
NDT & E International
0963-8695
Publisher
Elsevier
Publication Date
Volume
35
Issue
7
Identifiers
DOI: 10.1016/s0963-8695(02)00022-1
Keywords
  • Eddy-Current Probe
  • Back-Propagation Neural Network
  • Composite Material
  • Defect And Tilt Signals
Disciplines
  • Computer Science
  • Education

Abstract

Abstract The process of teaching and testing of back-propagation neural networks to characterize defect signals and tilt of eddy-current probe signals by superimposing different noise levels has been analyzed. Results of classification by clustering have been considered. Using numerical simulation for justification, it has been shown that clustering increases the probability of signal recognition.

There are no comments yet on this publication. Be the first to share your thoughts.