作者：殷鹰1, 谢罗峰2, 黄泰博3
作者单位：1. 四川省特种设备检验研究院，四川 成都 610061;
2. 四川大学机械工程学院，四川 成都 610065;
3. 四川大学锦城学院智能制造学院，四川 成都 611731
关键词：深度学习;无损检测;磁瓦;内部缺陷
摘要：
针对开发一套智能化磁瓦内部缺陷检测设备的需求，提出一种基于深度一维卷积网络的智能识别方法。该方法通过原始时域信号训练深度一维卷积网络，利用卷积网络逐层挖掘信号隐藏特征能力完成智能诊断。与传统方法相比，利用深度一维卷积网络能够摆脱对专家经验和信号处理知识的依赖，以其强大的自动提取特征能力完成磁瓦内部缺陷的智能诊断。在3种类型磁瓦数据上进行特征提取和缺陷识别，实验结果表明，该方法能够有效地从声音信号中提取缺陷特征和识别，结合开发的机械设备，能够满足磁瓦内部缺陷智能化检测的需求。
A deep learning method for magnetic tile internal defect inspection based on acoustic vibration
YIN Ying1, XIE Luofeng2, HUANG Taibo3
1. Sichuan Special Equipment Inspection Institute, Chengdu 610061, China;
2. School of Mechanical Engineering, Sichuan University, Chengdu 610065, China;
3. School of Intelligent Manufacturing, Sichuan University Jincheng College, Chengdu 611731, China
Abstract: In order to develop an intelligent magnetic tile internal defect detection equipment, an intelligent recognition method based on one-dimensional convolution network was proposed. The proposed one-dimensional convolution network could be directly trained with the original time-domain signal. It could extract the potential features layer by layer, and complete the intelligent diagnosis. Compared with the traditional method, the proposed network could get rid of the dependence on expert experience and signal processing knowledge, and could automatically diagnose the internal defects of magnetic tile by its powerful automatic feature extraction ability. The experiments were carried out on three types of magnetic tile. The experimental results showed that the proposed method could effectively extract defect features and identify the defects. Combined with the developed equipment, it could meet the requirements of intelligent inspection of magnetic tile internal defect.
Keywords: deep learning;nondestructive testing;magnetic tile;internal defect
2020, 46(3):32-38 收稿日期: 2019-05-10;收到修改稿日期: 2019-08-02
基金项目: 国家自然科学青年基金项目(51205265)；四川省科技计划项目（2018GZ0289）
作者简介: 殷鹰(1983-),男,四川成都市人,高级工程师,博士,研究方向为无损检测
参考文献
[1] 黄沁元, 殷鹰, 赵越, 等. 基于双谱分析的磁瓦内部缺陷音频检测方法[J]. 四川大学学报(工程科学版), 2014(5): 188-194
[2] 黄沁元, 殷鹰, 赵越, 等. 基于模糊聚类双谱的磁瓦内部缺陷无损检测方法[J]. 无损检测, 2014, 36(12): 15-19
[3] HUANG Q, YIN Y, YIN G. Automatic classification of magnetic tiles internal defects based on acoustic resonance analysis[J]. Mechanical systems and signal processing, 2015, 60-61: 45-58
[4] XIE L, YIN M, HUANG Q, et al. Internal defect inspection in magnetic tile by using acoustic resonance technology[J]. Journal of Sound and Vibration, 2016, 383: 108-123
[5] XIE L, HUANG Q, ZHAO Y, et al. Inspection of magnetic tile internal cracks based on impact acoustics[J]. Nondestructive Testing and Evaluation, 2015, 30(2): 147-164
[6] 赵越, 殷鸣, 黄沁元, 等. 基于WPT-ANN的磁瓦内部缺陷音频检测[J]. 中国测试, 2015,41(6): 81-85
[7] 谢罗峰, 徐慧宁, 黄沁元, 等. 应用双树复小波包和NCA-LSSVM检测磁瓦内部缺陷[J]. 浙江大学学报(工学版), 2017(1): 184-191
[8] HINTON G E. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786): 504-507
[9] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90
[10] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444
[11] 郭亮, 高宏力, 张一文, 等. 基于深度学习理论的轴承状态识别研究[J]. 振动与冲击, 2016(12): 166-170
[12] 雷亚国, 贾峰, 周昕, 等. 基于深度学习理论的机械装备大数据健康监测方法[J]. 机械工程学报, 2015(21): 49-56
[13] 赵光权, 葛强强, 刘小勇, 等. 基于DBN的故障特征提取及诊断方法研究[J]. 仪器仪表学报, 2016(9): 1946-1953
[14] TURKER I, SERKAN K, LEVENT E, et al. Real-time motor fault detection by 1-D convolutional neural networks[J]. IEEE Transactions on Industrial Electronics, 2016, 63(11): 7067-7075
[15] ZHANG W, LI C, PENG G, et al. A deep convolutional neural network with new training methods for bearing fault diagnosis under noisy environment and different working load[J]. Mechanical Systems and Signal Processing, 2018, 100: 439-453
[16] ZHANG W, PENG G, LI C, et al. A new deep learning model for fault diagnosis with good anti-noise and domain adaptation ability on raw vibration signals[J]. Sensors, 2017, 17(2): 425
[17] IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]// International Conference on International Conference on Machine Learning. JMLR.org, 2015.
