Abstract：Aiming at the problem that most of the existing time-domain modal parameter identification methods are difficult to set order and resist noise poorly, an unsupervised learning Convolution Neural Network (CNN) method for vibration signal modal identification was proposed. The proposed algorithm was improved on the basis of CNN. Firstly, the CNN applied to two-dimensional image processing was changed into the CNN to deal with one-dimensional signal. The input layer was changed into the vibration signal set of modal parameters to be extracted, and the intermediate layer was changed into several one-dimensional convolution layers, sampled layers, and output layer was the set of N-order modal parameters corresponding to the signal. Then, in the error evaluation, the network calculation result (N-order modal parameter set) was reconstructed by the vibration signals. Finally, the squared sum of the difference between the reconstructed signal and the input signal was taken as the network learning error, which makes the network become an unsupervised learning network, and avoids the ordering problem of modal parameter extraction algorithm. The experimental results show that when the constructed CNN is applied to modal parameter extraction, compared with the Stochastic Subspace Identification (SSI) algorithm and its Local Linear Embedding (LLE) algorithm, the convolutional neural network identification accuracy is higher than that of the SSI algorithm and the LLE algorithm under noise interference. It has strong noise resistance and avoids the ordering problem.
[1] 曹树谦,张文德,萧龙翔.振动结构模态分析理论、实验与应用[M].天津:天津大学出版社,2001:152-156.(CAO S Q, ZHANG W D, XIAO L X. The Theory, Experiment and Application of Structure Vibration Modal Analysis[M]. Tianjin:Tianjin University Press, 2001:152-156.) [2] 续秀忠,华宏星,陈兆能.基于环境激励的模态参数辨识方法综述[J].振动与冲击,2002,21(3):1-5.(XU X Z, HUA H X, CHEN Z N. Review of modal identification method based on ambient excitation[J]. Journal of Vibration and Shock, 2002, 21(3):1-5.) [3] 秦世强,蒲黔辉,施洲.环境激励下大型桥梁模态参数识别的一种方法[J].振动与冲击,2012,31(2):95-100.(QIN S Q, PU Q H, SHI Z. A method of modal parameters identification using ambient vibration measurements for a large-scale bridge[J]. Journal of Vibration and Shock, 2012, 31(2):95-100.) [4] 单德山,李乔.桥梁结构模态参数的时频域识别[J].桥梁建设,2015,31(2):26-31.(SHAN D S, LI Q, Model parameters of bridge structure in time-frequency domain[J]. Bridge Construction, 2015, 31(2):26-31.) [5] 付志超,程伟,李晶,等.一种基于神经网络的模态参数识别方法[J].机械强度,2010,32(6):899-904.(FU Z C,CHENG W,LI J,et al.Model parameters identification via an ANN-based method[J]. Journal of Mechanical Strength, 2010, 32(6):899-904.) [6] 白俊卿,闫桂荣,王成.利用局部线性嵌入的模态识别[J].西安交通大学学报,2013,47(1):85-89.(BAI J Q, YAN G R, WANG C. Modal identification method following locally linear embedding[J]. Journal of Xi'an Jiaotong University, 2013, 47(1):85-89. [7] 孙宗光,高赞明,倪一清.基于神经网络的损伤构件及损伤程度识别[J].工程力学,2006,23(2):18-22.(SUN Z G, GAO Z M, NI Y Q. Identification of damaged members and damage extent in bridge deck by neural network[J]. Engineering Mechanics, 2006, 23(2):18-22.) [8] 许斌,龚安苏,贺佳,等.基于神经网络模型的结构参数提取新方法[J].工程力学,2011,28(4):35-41.(XU B, GONG A S, HE J, et al. A novel structural parameter extraction with neural network based nonparametric model[J]. Engineering Mechanics, 2011, 28(4):35-41.) [9] 代煜,张建勋.基于小波变换和反向传播网络的模态参数辨识[J].振动与冲击,2012,31(3):55-59.(DAI Y, ZHANG J X. Identification of modal parameters based on wavelet transform and back propagation network[J]. Journal of Vibration and Shock, 2012, 31(3):55-59.) [10] 张晴晴,刘勇,潘接林,等.基于卷积神经网络的连续语音识别[J].工程科学学报,2015,37(9):1212-1217.(ZHANG Q Q, LIU Y, PAN J L, et al. Continuous speech recognition by convolutional neural networks[J]. Journal of University of Science and Technology Beijing, 2015, 37(9):1212-1217.) [11] DRUKI A A, SPITSYN V G. Development of sequential optimizational algorithms for object detection in images[C]//Proceedings of the 2015 International Siberian Conference on Control and Communications. Piscataway, NJ:IEEE, 2015:21-23. [12] 徐姗姗,刘应安,徐昇.基于卷积神经网络的木材缺陷识别[J].山东大学学报(工学版),2013,43(2):23-28.(XU S S, LIU Y A, XU S. Wood defects recognition based on the convolutional neural network[J]. Journal of Shandong University (Engineering Science), 2013, 43(2):23-28.) [13] 谢智歌,王岳青,窦勇,等.基于卷积-自动编码机的三维形状特征学习[J].计算机辅助设计与图形学学报,2015,27(11):2058-2064.(XIE Z G, WANG Y Q, DOU Y, et al. 3D feature learning via convolutional auto-encoder extreme learning machine[J]. Journal of Computer-Aided Design and Computer Graphics, 2015, 27(11):2058-2064.) [14] ABDEL-HAMID O, MOHAMED A, JIANG H, et al. Convolutional neural networks for speech recognition[J]. IEEE Transactions on Audio, Speech and Language Processing, 2014, 22(1):1533-1545. [15] SWIETOJANSKI P,GHOSHAL A,RENALS S. Convolutional neural networks for distant speech recognition[J]. IEEE Signal Processing Letters, 2014, 21(9):1120-1124. [16] BOUVRIE J. Notes on convolutional neural networks[EB/OL].[2016-01-25]. https://core.ac.uk/download/pdf/86960.pdf.
2017, Vol. 37 
