摘要
使用改进的功率谱函数对电磁超声缺陷信号进行了缺陷的定性分析,使用小波包能量谱对电磁超声缺陷接收信号进行了特征提取,从小波包的小波函数选取、分解层次及特征参数的噪声鲁棒性3个方面开展了讨论分析。结果表明:通过选择适当的小波函数和小波包分解层次,小波包能量谱的能量比可以精细地反映信号的特征;基于小波包能量谱的特征参数具有良好的损伤敏感性及噪声鲁棒性,能在强噪声影响下实现对EMAT不同损伤类型的判别。
This paper used the improved power spectrum to qualitative analysis the electromagnetic ultrasonic flaw signals, fol- lowed by the energy spectrum of wavelet packet distill the electromagnetic ultrasonic flaw received signal feature, discusing and analyzing the characteristic parameters from wavelet packet wavelet function selected, decomposition level and noise robustness. The results show that：through selectting the appropriate wavelet function and wavelet packet decomposition level, wavelet packet energy spectrum of energy ratio can subtly reflect the characteristics of the signal;The wavelet packet energy spectrum of characteristic parameters have good damnification-sensitive and noise-robustness,besides ,which can distinguish the EMAT different types of injury under the strong noise.
出处
《仪表技术与传感器》
CSCD
北大核心
2013年第2期79-82,共4页
Instrument Technique and Sensor
关键词
电磁超声
缺陷
小波包
特征提取
electromagnetic ultrasonic
defects
wavelet packet
feature extraction
