Deterministic methodology stabilizing with SVD and fuzzy inference for crack shapes on ECT by Laplace transform BEM
Deterministic methodology stabilizing with SVD and fuzzy inference for crack shapes on ECT by Laplace transform BEM
复制标题
使用 SVD 和拉普拉斯变换边界元法对 ECT 上的裂纹形状进行模糊推理来稳定确定性方法
DOI：
10.1109/20.767387
复制
发表时间：
1999
期刊：
影响因子：
--
通讯作者：
M. Enokizono
中科院分区：
文献类型：
--
作者：
Y. Tsuchida;K. Shibao;M. Enokizono
在本文中，描述了通过奇异值分解（SVD）和模糊推理来稳定求解矩阵逆问题的确定性方法。由于矩阵的稳定性可以通过奇异值导出的条件数得知，因此可以利用正则化参数或模糊推理来适当地稳定矩阵。将所提出的方法应用于涡流检测（ECT）裂纹形状识别，并讨论了其有效性。
In this paper, the deterministic methodology stabilizing with singular value decomposition (SVD) and fuzzy inference is described to solve matrices from inverse problems. Since the stability of the matrices can be known by the condition number derived from the singular value, the regularization parameter or fuzzy inference can be utilized to stabilize them properly. The proposed method is applied to crack shape identification on an eddy current testing (ECT) and the effectiveness of it is discussed.