GFresNet-2D
收藏资源简介：
A deep learning database and network for focusing guided wave defect detection Since the paper is being submitted, the database set will be published after the paper is accepted. Database set information:The defects are classified as three types and specimens with no defect are also included. In the established database set, the defect depth ranges from 10% to 50%, with 10% intervals. In addition, the radius of the pinhole defect ranges from 0.5 mm to 3 mm, and the sizes of the crack defect range from 1×5 mm2 to 2×10 mm2, and the sizes of the corrosion defect range from 5×5 mm2 to 10×10 mm2. Each defect contains 1500 signal data, and the ratio of the training, validation, and test data sets are divided into 6:2:2 in this work. The data storage format and explain the descriptive data (take the pinhole defect signal with a radius of 3 mm and a depth of 10% as an example). The data set has a total of 48,060,000 signal value data and contains detailed information about defects. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 ··· Data -1 1 0 0 3 10 0 0 0 -2 0 0 0 0 ··· Title: Development of frequency-mixed point-focusing SH guided wave EMAT for defect inspection using deep neural network Author: Hongyu Sun, Songling Huang, State Key of Power Systems, Department of Electrical Engineering, Tsinghua University, Beijing, 10084, China. Email: sunhy18@mails.tsinghua.edu.cn If you use our code and database set, please cite our paper [however, not published]. NOTICE: Reviewers can obtain the data set password from the end of the paper's Abstract to run the code. Development environment： TensorFlow 2.2 CUDA 10.1 Python 3.7
用于聚焦导波缺陷检测的深度学习数据库与网络。本论文目前处于投稿阶段，数据库集将在论文录用后正式发布。 数据库集说明：缺陷共分为三类，同时包含无缺陷试样。所构建的数据库集中，缺陷深度范围为10%至50%，步长为10%。其中，针孔缺陷的半径范围为0.5 mm至3 mm；裂纹缺陷的尺寸范围为1×5 mm²至2×10 mm²；腐蚀缺陷的尺寸范围为5×5 mm²至10×10 mm²。每类缺陷对应1500条信号数据，本研究中将训练集、验证集与测试集的划分比例设为6:2:2。下文将结合数据存储格式对描述性数据进行说明（以半径3 mm、深度10%的针孔缺陷信号为例）。本数据集总计包含48060000条信号值数据，并附带缺陷的详细相关信息。示例数据格式如下： 编号：1 2 3 4 5 6 7 8 9 10 11 12 13 14 ··· 信号值：-1 1 0 0 3 10 0 0 0 -2 0 0 0 0 ··· 论文标题：面向缺陷检测的混频点聚焦SH导波（Shear Horizontal Guided Wave）电磁超声换能器（Electro-Magnetic Acoustic Transducer, EMAT）的深度学习检测方法研发 作者：孙宏宇、黄松岭，清华大学电机工程系电力系统国家重点实验室，北京 10084，中国 邮箱：sunhy18@mails.tsinghua.edu.cn 若您使用本代码与数据库集，请引用本论文（目前尚未正式发表）。 注意：审稿人可从论文摘要末尾获取数据集密码以运行代码。 开发环境：TensorFlow 2.2、CUDA 10.1、Python 3.7
