基于经验模态分解的齿轮裂纹声发射检测  

Study on Acoustic Emission Gear Crack Detection Based on EMD Signal Analysis Method

作　　者：姚晓山[1,2] 张永祥[1] 明廷涛[1] 

机构地区：[1]海军工程大学船舶与动力学院,武汉430033 [2]空军雷达学院军通系,武汉430019

出　　处：《无损检测》2009年第6期464-467,共4页Nondestructive Testing

基　　金：湖北省自然科学基金资助项目(2006ABA011)

摘　　要：针对传感器检测到的齿轮裂纹声发射信号存在能量微弱且被噪声湮没的问题,首先将齿轮裂纹产生的声发射信号进行自适应噪声对消,消除环境噪声的干扰;然后将信号进行数字带通滤波,进一步抑制无用信号;最后将信号进行经验模态分解,提取具有裂纹冲击性质的固有模态分量。试验信号分析结果表明,该方法成功重构了齿轮裂纹声发射信号。A method was proposed to extract acoustic emission signal of gear crack, aiming at detecting fault signal mixed with great noise. Firstly, an adaptive noise compensator was designed to eliminate interference of background noise and therefore improve signal-to-noise ratio. Secondly, a digital bandpass filter was used to further suppress unnecessary signal. Lastly, the signal was decomposed into a finite intrinsic mode functions by empirical mode decomposition. Experiment demonstrated that this method could reconstruct acoustic emission signal of gear crack effectively.

关 键 词：声发射检测 自适应噪声对消 经验模态分解 带通滤波 齿轮 裂纹 

分 类 号：TG115.28[金属学及工艺—物理冶金]