Published April 9, 2005 | Version v1
Journal article Metadata-only

Reconstruction of Flaw Profiles Using Neural Networks and Multi-Frequency Eddy Current System

Creators

  • 1. Technical University of Szczecin, al. Piastow 19, 70-310 Szczecin (Poland)

Description

The objective of this paper is to identify profiles of flaws in conducting plates. To solve this problem, application of a multi-frequency eddy current system (MFES) and artificial neural networks is proposed. Dynamic feed-forward neural networks with various architectures are investigated. Extended experiments with all neural models are carried out in order to select the most promising configuration. Data utilized for the experiments were obtained from the measurements performed on the Inconel plates with EDM flaws

Additional details

Identifiers

DOI
10.1063/1.1916759;

Publishing Information

Journal Title
AIP Conference Proceedings
Journal Volume
760
Journal Issue
1
Journal Page Range
p. 820-827
ISSN
0094-243X
CODEN
APCPCS

Conference

Title
Conference on review of progress in quantitative nondestructive evaluation
Dates
25-30 Jul 2004
Place
Golden, CO (United States)

INIS

Country of Publication
United States
Country of Input or Organization
International Atomic Energy Agency (IAEA)
INIS RN
36095004
Subject category
S36: MATERIALS SCIENCE;
Resource subtype / Literary indicator
Conference
Descriptors DEI
CHROMIUM ALLOYS; DEFECTS; EDDY CURRENT TESTING; EDDY CURRENTS; NEURAL NETWORKS; NICKEL ALLOYS
Descriptors DEC
ALLOYS; CURRENTS; ELECTRIC CURRENTS; ELECTROMAGNETIC TESTING; MATERIALS TESTING; NONDESTRUCTIVE TESTING; TESTING; TRANSITION ELEMENT ALLOYS

Optional Information

Notes
(c) 2005 American Institute of Physics