摘要
在用超声脉冲回波法检测金属-非金属粘接结构的质量以及测量非金属层厚度时,需要提取出各界面回波间的延迟时间作为特征值。当非金属层较薄时,各回波信号间会产生严重混叠,难以准确获得时延估值。提出了一种多界面粘接结构的高分辨率时延估计方法,综合利用了维纳滤波解卷积技术和最大熵谱估计方法,从混叠信号中得到精确的时延估值,很好地解决了特征值提取困难的问题,并用模拟信号和实验信号验证了该方法。
In the ultrasonic testing of metal-nonmetal bonding structure quality using pulse echo technique,we need to extract the time delay eigenvalue between interface echoes to measure the thickness of nonmetal layers.When the nonmetal layers are quite thin,serious aliasing appears in the ultrasonic echo signals,which makes time delay estimation value inaccurate.Aiming at this problem,this paper brings forward a time delay estimation method of high resolution basing on multi-interface bonding structure.The method utilizes both wiener filter deconvolution technique and maximum entropy spectrum estimation to get accurate time delay value from aliasing signals,which resolves problems in the eigenvalue extraction very well.The estimation results are validated by analog signals and experiment signals.
出处
《材料工程》
EI
CAS
CSCD
北大核心
2006年第z1期389-392,共4页
Journal of Materials Engineering
基金
国家自然科学基金项目资助(50375046)
关键词
金属-非金属粘接结构
超声检测
时延估计
最大熵谱估计
metal-nonmetal bonding structure
ultrasonic testing
time delay estimation
maximum entropy spectrum estimation