摘要
黄土振动压实的压应力信号是一个含噪的非平稳瞬态突变信号.为了分析信号的联合时频特性,提出基于小波变换去噪和奇异值检测的信号分析方法.利用多分辨小波分解,选择性的提取有用频段信号并抑制某些高频段噪声,可以有效实现信噪分离.利用小波变换奇异值分析方法,通过检测小波高频系数时序信号的2种模极大值,可以获得压应力开始出现时刻、抵达峰值时刻点,从而可计算出压应力从出现到抵达峰值的历经时间,这对于进一步研究压应力在土体中的分布和传播规律具有重要意义.
The stress signal of soil during vibrating compaction is unstable and transient saltation signals accompanied by strong noise. To study the joint time-frequency property of the soil stress signal during vibrating compaction,wavelet transform method for de-noising and detection is proposed. By multiresolution wavelet decomposition,the reconstructed signal keeps useful frequency band but suppress some high frequency band. That separates the noise from the signal efficiently. For singularity analysis,the time of the stress appearance and the time of the stress peak arrival are obtained by detecting two kinds of the modulus maximum of the wavelet transform coefficients. Thus the impaction time from the stress appearance to peak arrival can be calculated. It is of great importance for analyzing the distribution and the propagation law of the stress in the soil.
出处
《微电子学与计算机》
CSCD
北大核心
2010年第5期206-208,共3页
Microelectronics & Computer
基金
国家自然科学基金项目(60801057
50578017)
关键词
压应力
信号
小波变换
去噪
奇异值检测
stress
signal
wavelet transform
de-noising
singularity detection
