摘要
根据超声探伤信号的多小波阈值去噪原理,针对硬阈值函数不连续和软阈函数存在恒定偏差的缺点,提出了一种改进的软阈值函数。针对目前LabVIEW没有对多小波算法支持的缺点和MATLAB强大的数值处理能力,利用LabVIEW和MATLAB混合编程技术,在LabVIEW中实现了超声探伤信号的多小波阈值去噪处理,并显示消噪波形、保存消噪数据和频谱分析,扩展了LabVIEW的多小波信号去噪功能。实验结果表明,多小波阈值去噪与单小波阈值相比,在超声探伤信号中起到更好的去噪效果,改进的软阈值函数比传统软阈值函数明显地提高了信噪比,能够更有效地提取缺陷信息。
An improved soft-threshold function has been proposed to overcome the discontinuous shortcoming of the hardthreshold function and the invariable dispersion of soft-threshold function according to the denoising principle of ultrasonic detection signal by multi-wavelet. Because of the short comings of multi-wavelet algorithm in LabVIEW and powerful numerical capability of MATLAB, utilizing hybrid program by LabVIEW and MATLAB, a virtual de-noising instrument has been designed to realize the denoising of ultrasonic detection signal. Meanwhile, it can display the waveforms, save denoising data and the analysis of spectrum. The experiment result shows, by contrasting the denoising effect of wavelets and multi-wavelets, a better effect can be obtained by multi-wavelets. Compared with soft-threshold and hard-threshold; the improved soft-threshold function can improve SNR at a higher level and can also fetch flaw information more effectively.
出处
《微型机与应用》
2012年第5期75-78,83,共5页
Microcomputer & Its Applications
