摘要
鉴于多层金属结构的超声检测信号夹杂了大量的噪声,湮没了有用的裂纹缺陷信息,在对信号作多尺度分析的基础上,构造了Morlet-P小波并以此对信号进行了各种分析。当取多个尺度参数时,基于Morlet-P小波变换得到的小波系数都包含了缺陷信号的频率成分,解决了多分辨率分析时多尺度小波系数在分析窄频带缺陷信号时的不足。
The ultrasonic testing signal of multi-layered metal structure includes a mass of noise, and the noise submerges completely the useful flaw signal. Morlet-P wavelet was constructed during the multi-scale analysis. When more than one scale parameter was applied to the continuous wavelet transform based on the Morlet-P wavelet, the wavelet transform coefficient still included the flaw characters. So the shortcoming of multi-resolution analysis on the narrowband signal was overcome.
出处
《无损检测》
北大核心
2007年第6期319-321,共3页
Nondestructive Testing
基金
国家自然科学基金资助项目(大型复杂设备中金属基复合材料结构缺陷检测与诊断原理研究)
关键词
超声检测
多层金属结构
时频分析
多分辨率分析
信号处理
Ultrasonic testing
Multi-layered metal structure
Time-frequency analysis
Multi-resolution
Signalprocessing
