作者：李伟, 姜智通, 张璐莹, 蒋鹏

作者单位：东北石油大学机械科学与工程学院, 黑龙江 大庆 163318

关键词：碳纤维复合材料;声发射;纤维断裂;经验模态分解;主成分分析;支持向量机

摘要：

针对碳纤维复合材料层合板剪切过程中所产生的纤维断裂及基体开裂声发射信号的数据样本数量多、分布随机、变化形式较为离散等问题，提出一种可用于识别纤维断裂及基体开裂两种损伤类型的方法。首先，利用经验模态分解(EMD)对纤维断裂及基体开裂的声发射信号进行时频变换；然后，对分解后信号进行快速傅里叶变换(FFT)以获得特征频率集，再利用主成分分析法(PCA)对特征频率集进行降维处理；最后，利用支持向量机(SVM)实现纤维断裂及基体开裂信号进行损伤模式识别。结果表明，此方法可较为准确地识别纤维断裂及基体开裂两种信号。针对碳纤维复合材料层合板剪切过程所产生的声发射信号，模型的总识别率达85.8%。

Pattern identification method for acoustic emission signals of damage in carbon fiber reinforced polymer
LI Wei, JIANG Zhitong, ZHANG Luying, JIANG Peng
School of Mechanical Science and Engineering, Northeast Petroleum University, Daqing 163318, China
Abstract: Aiming at the problems of large number of data samples, random distribution and discrete variation of acoustic emission signals of fiber fracture and matrix cracking in carbon fiber composite laminates during shearing process, a method for two types of injury of identifying fiber breakage and matrix cracking is proposed. Firstly, adopting empirical mode decomposition (EMD) for time-frequency transform of the acoustic emission signals of fiber fracture and matrix cracking. Then, fast Fourier transform (FFT) is applied to the decomposed signal to obtain the characteristic frequency set, and principal component analysis (PCA) is used to reduce the dimension of the characteristic frequency set. Finally, the damage pattern recognition of fiber fracture and matrix cracking signal is realized by support vector machine (SVM). The results indicate that this method can identify fiber fracture and matrix cracking signals accurately. The total recognition rate of the model is 85.8% specific to the acoustic emission signal generated during shearing of carbon fiber composite laminates.
Keywords: carbon fiber reinforced polymer;acoustic emission;fiber breakage;empirical mode decomposition;principal component analysis;support vector machine
2020, 46(6):121-128  收稿日期: 2019-12-17;收到修改稿日期: 2020-01-15
基金项目: 国家重点研发计划(2017YFC0805600)
作者简介: 李伟(1970-),男,黑龙江大庆市人,教授,博士,主要从事现代无损检测技术研究
参考文献
[1] 李伟, 李英年, 蒋鹏, 等. T300碳纤维复合材料损伤声发射信号的有监督模式识别[J]. 无损检测, 2016, 38(2): 9-13
[2] 黄振峰, 刘永坚, 毛汉颖, 等. 基于K熵和关联维数的金属疲劳损伤过程的声发射信号特征分析[J]. 振动与冲击, 2017, 36(15): 210-214
[3] SURGEON M, VANS W E, WEVERS M, et al. Acoustic emission during tensile testing of SiC-fibre-reinforced BMAS glass-ceramic composites[J]. Composites Part A, 1997, 28(5): 473-480
[4] 金周庚, 刘哲军, 王健, 等. B/Al复合材料变形和断裂过程声发射特性[J]. 稀有金属, 1999(3): 2-7
[5] GREGORY N,MORSCHER. Modal acoustic emission of damage accumulation in a woven SiC/SiC composite[J]. Composites Science and Technology, 1999, 59(5): 687-697
[6] RABIEI A, ENOKI M, KISHI T. A study on fracture behavior of particle reinforced metal matrix composites by using acoustic emission source characterization[J]. Materials Science & Engineering A, 2000, 293(1): 81-87
[7] 刘怀喜, 张恒, 闫耀辰. 人工神经网络在碳/环氧复合材料声发射检测中的应用[J]. 无损检测, 2004(12): 628-632
[8] MAREC A, THOMAS J H, GUERJOUMA R E. Damage characterization of polymer-based composite materials: Multivariable analysis and wavelet transform for clustering acoustic emission data[J]. Mechanical Systems and Signal Processing, 2007, 22(6): 1441-1464
[9] 李海斌, 阳建红, 刘承武, 等. 复合材料随机渐进失效分析与声发射监测[J]. 复合材料学报, 2011, 28(1): 223-229
[10] 栗丽. 基于声发射信号分析的2D及3D纺织结构复合材料损伤机制研究[D]. 上海: 东华大学, 2015.
[11] EMMANUEL M, CHRISTOPHER B, GREGORY N M, et al. Feasibility and limitations of damage identification in composite materials using acoustic emission[J]. Composites Part A, 2015, 75: 77-83
[12] 马云阔, 戴光, 李伟,等. 基于小波包特征熵的碳纤维复合材料损伤声发射信号特征提取方法[J]. 无损检测, 2017, 39(5): 76-80
[13] JALAL Y, MEHDI A N, HOSSEIN H T, et al. Damage evaluation of laminated composite material using a new acoustic emission Lamb-based and finite element techniques[J]. Applied Composite Materials, 2018, 25(5): 1021-1040
[14] 方赛银, 邱荣祖, 李明. 基于改进EMD算法的木材声发射信号特征研究[J]. 振动与冲击, 2018, 37(23): 292-298
[15] 张林燕, 骆红云. 基于声发射信号的16Mn钢损伤分析[J]. 金属热处理, 2019, 44(S1): 608-610
[16] 孟超, 张熙, 何志祥, 等. 基于声发射的碳纤维复合材料拉伸损伤特性研究[J]. 玻璃钢/复合材料, 2019(10): 5-9
[17] 袁忠, 黄频波, 耿文霞. 基于近似熵谱的碳纤维复合材料层压板拉伸损伤声发射分析[J]. 宇航材料工艺, 2014, 44(4): 73-77
[18] HUANG N E, SHEN Z, LONG S,et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 1998, 454: 1971
[19] DENG Y J, WANG W, QIAN C C, et al. Boundary-processing-technique in EMD method and Hilbert transform[J]. Chinese Science Bulletin, 2001(11): 954-961
[20] HERVÉ A, WILLIAMS L J. Principal component analysis[J]. Wiley Interdisciplinary Reviews: Computational Statistics, 2010, 2(4): 37-52
[21] JAHAN T, MEHDI A N. Classification of acoustic emission signals collected during tensile tests on unidirectional ultra high molecular weight polypropylene fiber reinforced epoxy composites using principal component analysis[J]. Russian Journal of Nondestructive Testing, 2011, 47(7): 491
[22] CHRISTOPHER J C B. A tutorial on support vector machines for pattern recognition[J]. Data Mining and Knowledge Discovery, 1998, 2(2): 121-167
[23] SUYKENS J A K, VANDEWALLE J. Least squares support vector machine classifiers[J]. Neural Processing Letters, 1999, 9(3): 293-300
[24] 张学工. 关于统计学习理论与支持向量机[J]. 自动化学报, 2000(1): 36-46