Answers are grounded in this paper and cite their sources.
Quantitative non-destructive evaluation, especially sizing of piping wall thinning in nuclear power plants is still a difficult and urgent issue. In this paper, an inversion approach for PECT (pulsed eddy current testing) signals is developed based on ANN (artificial neural network) method at first for profile reconstruction of wall thinning, the sizing result of NN is then utilized as the initial value of the CG (conjugate gradient) inversion scheme to overcome the shortages of both the NN (accuracy problem) and CG (local minimum problem) methods. Several reconstruction examples using the proposed hybrid strategy indicate that the combination of NN and CG methods is rather effective for wall thinning reconstruction from PECT signals in view of both the robustness and sizing accuracy.
No takes yet. Share an insight, caveat, or question.
Xie et al. (2013) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: