卷积神经网络的缺陷类型识别分析
Recognition and analysis of defect types by convolutional neural network
查看参考文献23篇
|
文摘
|
该文提出一种基于卷积神经网络直接对阵列超声检测原始信号进行缺陷类型识别的方法,该方法无需对超声回波原始信号进行特征提取。文章研究对比了不同卷积神经网络及其优化的识别性能。首先采用超声相控阵系统对不同试块上的平底孔、球底孔、通孔三种缺陷进行超声检测,然后利用LeNet5、VGG16和ResNet三种卷积神经网络对一维和二维数据分别进行缺陷识别,并使用Leaky ReLU、Dropout、Batch Normalization等来对网络进行优化,对比分析识别准确率和效率等。研究结果表明,虽然一维卷积在训练速度上快于二维,但是二维卷积在识别准确度方面更高,同时网络模型结构如果过于复杂会导致准确率的下降,数据增强和优化方式有助于加快收敛速度、提高准确率。 |
|
其他语种文摘
|
In this paper, a method of the defect type recognition based on the convolutional neural network is proposed by the original signal from the array ultrasonic testing. The method does not require feature extraction of the original ultrasonic echo signals. The recognition performance and optimization characteristic of different convolutional neural networks are studied and analyzed. Firstly, the experiments are conducted by the ultrasonic phased array system to collect detection signals which comes from the flat-bottom hole, spherical-bottom hole and via hole. Then, Lenet5, VGG16 and Resnet convolutional neural networks are used to identify the defects in one and two dimensions, and the network is optimized by Leaky ReLU, Dropout and Batch Normalization respectively. Subsequently, the identification accuracy and efficiency are analyzed. It is shown that 2-d convolution has higher recognition accuracy although its training speed is slower than that of 1-d convolution. At the same time, the identification accuracy will decline if the network model structure is too complex, and the data enhancement and optimization methods can help to accelerate the convergence speed and improve the accuracy. |
|
来源
|
应用声学
,2022,41(2):301-309 【扩展库】
|
|
DOI
|
10.11684/j.issn.1000-310X.2022.02.017
|
|
关键词
|
卷积神经网络
;
超声检测
;
缺陷类型识别
|
|
地址
|
1.
中国科学院大学, 北京, 100049
2.
中国科学院声学研究所, 声场声信息国家重点实验室, 北京, 100190
3.
中国科学院力学研究所, 非线性力学国家重点实验室, 北京, 100190
4.
中国科学院大学工程科学学院, 北京, 100049
|
|
语种
|
中文 |
|
文献类型
|
研究性论文 |
|
ISSN
|
1000-310X |
|
学科
|
物理学 |
|
基金
|
中国科学院声学研究所青年英才计划项目
;
中国科学院基础前沿科学研究计划
;
非线性力学国家重点实验室开放基金资助项目
;
中国科学院声学研究所声场声信息国家重点实验室基金
|
|
文献收藏号
|
CSCD:7168291
|
参考文献 共
23
共2页
|
|
1.
彭应秋.
超声波检测工艺,1990:5-7
|
CSCD被引
2
次
|
|
|
|
|
2.
Song S J. Ultrasonic flaw classification in weldments using probabilistic neural networks.
Journal of Nondestructive Evaluation,1992,11(2):69-77
|
CSCD被引
2
次
|
|
|
|
|
3.
Masnata A. Neural network classification of flaws detected by ultrasonic means.
NDT & E International,1996,29(2):87-93
|
CSCD被引
7
次
|
|
|
|
|
4.
Margrave F W. The use of neural networks in ultrasonic flaw detection.
Measurement,1999,25(2):143-154
|
CSCD被引
1
次
|
|
|
|
|
5.
卢超. 小波分析和人工神经网络在金属超声无损检测缺陷分类中的应用.
南昌航空工业学院学报,2001,15(3):51-54
|
CSCD被引
2
次
|
|
|
|
|
6.
Liu S W. Detection of cracks using neural networks and computational mechanics.
Computer Methods in Applied Mechanics and Engineering,2002,191(25/26):2831-2845
|
CSCD被引
6
次
|
|
|
|
|
7.
Drai R. Time frequency and wavelet transform applied to selected problems in ultrasonics NDE.
NDT & E International,2002,35(8):567-572
|
CSCD被引
6
次
|
|
|
|
|
8.
Veiga J L B C. The use of artificial neural network in the classification of pulse-echo and TOFD ultra-sonic signals.
Journal of the Brazilian Society of Mechanical Sciences and Engineering,2005,27(4):394-398
|
CSCD被引
1
次
|
|
|
|
|
9.
Sambath S. Automatic defect classification in ultrasonic NDT using artificial intelligence.
Journal of Nondestructive Evaluation,2011,30(1):20-28
|
CSCD被引
9
次
|
|
|
|
|
10.
Filho E F S. Flexible decision support system for ultrasound evaluation of fiber-metal laminates implemented in a DSP.
NDT & E International,2016,79:38-45
|
CSCD被引
2
次
|
|
|
|
|
11.
Cruz F C. Efficient feature selection for neural network based detection of flaws in steel welded joints using ultrasound testing.
Ultrasonics,2017,73:1-8
|
CSCD被引
9
次
|
|
|
|
|
12.
Lecun Y. Gradient-based learning applied to document recognition.
Proceedings of the IEEE,1998,86:2278-2324
|
CSCD被引
2118
次
|
|
|
|
|
13.
Krizhevsky A. ImageNet classification with deep convolutional neural networks.
Communications of The ACM,2017,60(6):84-90
|
CSCD被引
3075
次
|
|
|
|
|
14.
施成龙. 利用深度神经网络和小波包变换进行缺陷类型分析.
声学学报,2016,41(4):499-506
|
CSCD被引
10
次
|
|
|
|
|
15.
Meng M. Ultrasonic signal classification and imaging system for composite materials via deep convolutional neural networks.
Neurocomputing,2017,257:128-135
|
CSCD被引
15
次
|
|
|
|
|
16.
Khumaidi A.
Welding defect classification based on convolution neural network (CNN) and Gaussian kernel,2017:261-265
|
CSCD被引
1
次
|
|
|
|
|
17.
Munir N. Investigation of deep neural network with drop out for ultrasonic flaw classification in weldments.
Journal of Mechanical Science and Technology,2018,32(7):3073-3080
|
CSCD被引
1
次
|
|
|
|
|
18.
Munir N. Convolutional neural network for ultrasonic weldment flaw classification in noisy conditions.
Ultrasonics,2019,94:74-81
|
CSCD被引
4
次
|
|
|
|
|
19.
张重远. 基于相似矩阵盲源分离与卷积神经网络的局部放电超声信号深度学习模式识别方法.
电网技术,2019,43(6):1900-1907
|
CSCD被引
26
次
|
|
|
|
|
20.
Munir N. Performance enhancement of convolutional neural network for ultrasonic flaw classification by adopting autoencoder.
NDT & E International,2020,111:102218
|
CSCD被引
2
次
|
|
|
|
|
了解气象卫星更多信息，请访问