基于时序二维化和注意力机制的法兰螺栓连接状态识别

张洪1,2,刘彬彬1,2,李云飞3

振动与冲击 ›› 2022, Vol. 41 ›› Issue (9) : 181-188.

PDF(2651 KB)
PDF(2651 KB)
振动与冲击 ›› 2022, Vol. 41 ›› Issue (9) : 181-188.
论文

基于时序二维化和注意力机制的法兰螺栓连接状态识别

  • 张洪1,2,刘彬彬1,2,李云飞3
作者信息 +

Flange bolted connection state recognition based on time series two dimensionalization and attention mechanism

  • ZHANG Hong1,2, LIU Binbin1,2, LI Yunfei3
Author information +
文章历史 +

摘要

针对采用振动对法兰螺栓进行动态激励时,采集得到的声发射信号难以对螺栓连接状态进行精确识别的问题。提出了一种基于注意力网络的信号识别分类方法;采用分段聚合近似和格拉姆角场将声发射信号编码为二维图像,用图像代替一维信号作为诊断依据。为提高对图像的特征提取识别能力,引入注意力机制到残差神经网络中,设计了分组卷积诊断模型,实现了法兰螺栓连接状态的高精度识别。对比所选多种方法的试验结果表明,本文方法可以有效地识别螺栓的连接状态,具有较强的细节特征提取能力和泛化性能。

Abstract

When the flange bolt connection is dynamically excited with vibration, it is difficult to identify the connection state of the flange bolt with the collected acoustic emission signal. In order to solve this problem, a signal recognition and classification method based on attention network is proposed, in which the AE signal is encoded into a two-dimensional image with piecewise aggregation approximation and Gramm angle field, and the image is used instead of one-dimensional signal as the basis for diagnosis. The attention mechanism is introduced into the residual neural network, for improving the ability of feature extraction and recognition of the image. A group convolution diagnosis model is designed to realize the high-precision recognition of the flange bolt connection state. Compared with the selected methods, the experimental results show that this method can effectively identify the connection state of bolts and has strong detail feature extraction ability and generalization performance.

关键词

法兰螺栓 / 动态激励 / 声发射信号 / 格拉姆角场 / 注意力机制

Key words

Flange bolt / Dynamic excitation / Acoustic emission signal / Gramian angular field / Attention mechanism

引用本文

导出引用
张洪1,2,刘彬彬1,2,李云飞3. 基于时序二维化和注意力机制的法兰螺栓连接状态识别[J]. 振动与冲击, 2022, 41(9): 181-188
ZHANG Hong1,2, LIU Binbin1,2, LI Yunfei3. Flange bolted connection state recognition based on time series two dimensionalization and attention mechanism[J]. Journal of Vibration and Shock, 2022, 41(9): 181-188

参考文献

[1]  Wang Y Q, Wu J K, Liu H B, et al. Analysis of elastic interaction stiffness and its effect on bolt preloading[J]. International Journal of Mechanical Sciences, 2017, 130: 307-314.
[2]  王崴, 徐浩, 马跃, 等. 振动工况下螺栓连接自松弛机理研究[J]. 振动与冲击, 2014(22): 198-202.
WANG Wei, XU Hao, MA Yue, et al. Self-loosening mechanism of bolted joints under vibration [J]. Journal of Vibration and Shock, 2014(22): 198-202.
[3]  Wang F, Ho S C M, Song G. Monitoring of early looseness of multi-bolt connection: a new entropy-based active sensing method without saturation[J]. Smart Materials and Structures, 2019, 28(10): 10LT01.
[4]  艾延廷, 刘成明, 王志, 等. 基于非线性阻尼识别的螺栓连接检测技术[J]. 振动与冲击, 2020, 39(9): 138-143,180.
AI Yan-ting, LIU Cheng-ming, WANG Zhi, et al. Technology for detecting bolted joints based on nonlinear damping identification [J]. Journal of Vibration and Shock, 2020, 39(9): 138-143,180.
[5]  Shah J K , Mukherjee A . Monitoring and Imaging of Bolted Steel Plate Joints Using Ultrasonic Guided Waves[J]. Journal of Nondestructive Evaluation Diagnostics and Prognostics of Engineering Systems, 2020, 4(1):1-20.
[6]  杜飞, 张子涵, 徐超. 法兰螺栓松动的超声导波监测方法[J]. 压电与声光, 2019, 41(5): 679-684.
DU Fei, ZHANG Zi-han, XU Chao. Ultrasonic Guided Wave Monitoring Method for Flange Bolt Loosening [J]. Piezoelectrics & Acoustooptics, 2019, 41(5): 679-684.
[7]  潘勤学, 邵唱, 肖定国, 等. 基于形状因子的螺栓紧固力超声检测方法研究[J]. 兵工学报, 2019, 40(4): 880-888.
PAN Qin-xue, SHAO Chang, XIAO Ding-guo, et al. Study of Ultrasonic Measurement Method for Bolt Fastening Force Based on Shape Factor [J]. Acta Armamentarii, 2019, 40(4): 880-888.
[8]  孙朝明, 孙凯华, 孙鹏飞, 等. 超声法测量螺栓轴向力的测量准确度分析[J]. 振动与冲击, 2021, 40(8): 85-91.
SUN Chao-ming,SUN Kai-hua,SUN Peng-fei,et al. Analysis on measurement accuracy of bolt axial force by ultrasonic technique. Journal of Vibration and Shock, 2021, 40(8): 85-91.
[9]  Kong Q, Zhu J, Ho S C M, et al. Tapping and listening: A new approach to bolt looseness monitoring[J]. Smart Materials and Structures, 2018, 27(7): 07LT02.
[10]  Amerini F, Meo M. Structural health monitoring of bolted joints using linear and nonlinear acoustic/ultrasound methods[J]. Structural health monitoring, 2011, 10(6): 659-672.
[11]  王怡, 王宁, 卢萍, 等. 基于声发射原理的螺栓连接状态辨识方法研究[J]. 声学技术, 2010, 29(5): 453-456.
WANG Yi, WANG Ning, LU Ping, et al. The study of identifying the state of bolted joint structure based on acoustic emission principle [J]. Technical Acoustics, 2010, 29(5): 453-456.
[12]  张健奎, 王宁, 卢萍, 等. 辨识振动环境中两点螺栓连接状态的REE声发射指标[J]. 振动与冲击, 2013, 32(8): 179-182.
ZHANG Jian-kui, WANG Ning, LU Ping, et al. AE REE index to identify connecting state of a 2-bolt connected structure [J]. Journal of Vibration and Shock, 2013, 32(8): 179-182.
[13]  Wang Z, Oates T. Imaging time-series to improve classification and imputation[J]. arXiv preprint arXiv:1506.00327, 2015.
[14]  He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]. 2016 IEEE conference on computer vision and pattern recognition. Las Vegas: IEEE ,2016.
[15]  Mnih V, Heess N, Graves A. Recurrent models of visual attention[C]. Advances in neural information processing systems, models of visual attention[C]. Advances in neural information processing systems, 2014: 2204-2212.
[16]  Hu J, Shen L, Sun G. Squeeze-and-excitation networks[C]. Proceedings of the IEEE conference on computer vision and pattern recognition, 2018: 7132-7141.
[17]  Miao R, Shen R, Zhang S, et al. A Review of Bolt Tightening Force Measurement and Loosening Detection[J]. Sensors, 2020, 20(11): 3165.
[18]  Zhang Z, Liu M, Su Z, et al. Quantitative evaluation of residual torque of a loose bolt based on wave energy dissipation and vibro-acoustic modulation: A comparative study[J]. Journal of Sound and Vibration, 2016, 383: 156-170

PDF(2651 KB)

209

Accesses

0

Citation

Detail

段落导航
相关文章

/