首页> 外文会议>Annual Symposium on Quantitative Nondestructive Evaluation >CRACK ANGLE AND DEPTH ESTIMATION USING WAVELET PREPROCESSED NEURAL NETWORK
【24h】

CRACK ANGLE AND DEPTH ESTIMATION USING WAVELET PREPROCESSED NEURAL NETWORK

机译:使用小波预处理神经网络的裂缝角度和深度估计

获取原文
获取外文期刊封面目录资料

摘要

Crack parameters Estimator, based on MFL data analysis including Wavelet Transform on preprocessing step, are proposed. The 2D case was used to obtain only preliminary results and to demonstrate the feasibility of the approach. A further practical application to the magnetic inspection requires three-dimensional (3D) numerical models to construct an appropriate learning database for a NN classifier. To evaluate the generalization abilities of the NN classifier more extensively, a larger number of numerically obtained data is required. There is a further limitation to this analysis, namely that it have been concerned with modeling the crack without consideration being given to the magnetic behavior of the complete magnetizer. Future work includes more realistic crack shape, as well consideration of noise and vibration.
机译:提出了基于包括在预处理步骤的小波变换的MFL数据分析的裂缝参数估计器。 2D案例用于仅获得初步结果,并证明该方法的可行性。 对磁检查的进一步实际应用需要三维(3D)数值模型来构建适当的学习数据库,用于NN分类器。 为了更广泛地评估NN分级器的泛化能力,需要更大数量的数量的数据。 对该分析有一个进一步的限制,即它已经涉及建模裂缝而不考虑完整磁化器的磁性行为。 未来的工作包括更逼真的裂缝形状,同时考虑噪音和振动。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号