基于经验模态分解和小波变换声发射信号去噪  

De-nosing of acoustic emission signals based on empirical mode decomposition and wavelet transform

作　　者：于金涛[1,2] 赵树延[1] 王祁[1] 

机构地区：[1]哈尔滨工业大学自动化测试与控制系,哈尔滨150001 [2]哈尔滨商业大学计算机信息与工程学院,哈尔滨150028

出　　处：《哈尔滨工业大学学报》2011年第10期88-92,共5页Journal of Harbin Institute of Technology

基　　金：国家自然科学基金资助项目(60901042);黑龙江省自然科学基金资助项目(F201018)

摘　　要：为了解决声发射信号去噪问题,在分析经验模态分解去噪和小波阈值去噪优缺点的基础上,提出将二者相结合的去噪方法,包括IMF-Wavelet方法,EMD-Wavelet方法和Wavelet-EMD方法.利用标准信号及断铅模拟声发射信号对所研究方法进行了去噪性能分析.结果表明:对于标准信号,Wavelet-EMD方法无论在高信噪比还是低信噪比情况下,都具有较稳定的去噪效果;小波阈值去噪对声发射信号去噪效果不理想,EMD-Wavelet和Wavelet-EMD方法对于声发射信号去噪效果较稳定.To solve the acoustic emission （AE） signal de-noising problem, de-noising approaches based on empirical mode decomposition and wavelet transform were proposed, including IMF-Wavelet method, EMD- Wavelet method and Wavelet-EMD method. The standard noise signals and AE signals by pencil lead break were used to analyze the de-noising performance. The simulation results indicate that for the standard noise signals, Wavelet-EMD method has stable de-noising performance whether in high or low signal to noise ratio （SNR） case. Wavelet threshold method is not suitable for AE signal de-noising, EMD-Wavelet and Wavelet- EMD method have stable de-noising performance for AE signals.

关 键 词：经验模态分解 小波变换 阈值 去噪 声发射 

分 类 号：TG115[金属学及工艺—物理冶金]