Efficient Global Sensitivity Analysis of Model-Based Ultrasonic Nondestructive Testing Systems Using Machine Learning and Sobol’ Indices

Efficient Global Sensitivity Analysis of Model-Based Ultrasonic Nondestructive Testing Systems Using Machine Learning and Sobol’ Indices
复制标题

使用机器学习和 Sobol™ 指数对基于模型的超声无损检测系统进行高效的全局灵敏度分析

DOI:
10.1115/1.4051100
复制
发表时间:
2021
期刊:
Diagnostics and Prognostics of Engineering Systems
影响因子:
--
通讯作者:
Leifsson, Leifur
Leifsson, Leifur
中科院分区:
--
文献类型:
--
作者:
Nagawkar, Jethro;Leifsson, Leifur

文献摘要

参考文献

相似文献

这项工作的目的是通过用机器学习(ML)算法代替精确的基于物理的模型来降低对超声无损检测系统进行基于模型的灵敏度分析的成本,并快速计算Sobol指数。在这项工作中考虑的ML算法是神经网络(NN),卷积NN(CNN)和深度高斯过程(DGP)。这些算法的性能是通过固定数量的测试点上的均方根误差和达到目标精度所需的高保真样本数量来衡量的。算法进行了比较,在三个超声检测基准情况下,三个不确定性参数,即球形空洞缺陷下的聚焦和平面换能器和球形夹杂缺陷下的聚焦换能器。结果表明,三种情况下,NN分别需要35、100和35个样本。CNN分别需要35、100和56,而DGP分别需要84、84和56。
The objective of this work is to reduce the cost of performing model-based sensitivity analysis for ultrasonic nondestructive testing systems by replacing the accurate physics-based model with machine learning (ML) algorithms and quickly compute Sobol’ indices. The ML algorithms considered in this work are neural networks (NNs), convolutional NN (CNNs), and deep Gaussian processes (DGPs). The performance of these algorithms is measured by the root mean-squared error on a fixed number of testing points and by the number of high-fidelity samples required to reach a target accuracy. The algorithms are compared on three ultrasonic testing benchmark cases with three uncertainty parameters, namely, spherically void defect under a focused and a planar transducer and spherical-inclusion defect under a focused transducer. The results show that NNs required 35, 100, and 35 samples for the three cases, respectively. CNNs required 35, 100, and 56, respectively, while DGPs required 84, 84, and 56, respectively.
数据库生成和开发用于高效、深入的模拟研究
DOI: 10.1063/1.4940632
发表时间: 2016
影响因子: 0.6
作者:
R. Miorelli;X. Artusi;A. B. Abdessalem;C. Reboud
通讯作者: C. Reboud
通过射线追踪对对接接头超声检测进行建模
DOI: 10.1007/bf03185104
发表时间: 2001
期刊: KSME International Journal
影响因子: --
作者:
Y. Nam
通讯作者: Y. Nam
DOI: 10.1016/j.proeng.2014.11.059
发表时间: 2014
期刊: Procedia Engineering
影响因子: --
作者:
S. Subair;K. Balasubramaniam;P. Rajagopal;Anish Kumar;B. Rao;T. Jayakumar
通讯作者: T. Jayakumar
使用超声波测试评估铝制 FSW 接头
DOI: 10.1515/amm-2017-0353
发表时间: 2017
影响因子: 0.6
作者:
K. Adamus;P. Lacki
通讯作者: P. Lacki
涡流测试问题的克里金法
DOI: 10.1109/tmag.2010.2043418
发表时间: 2010
影响因子: 2.1
作者:
S. Bilicz;E. Vázquez;S. Gyimóthy;J. Pávó;M. Lambert
通讯作者: M. Lambert