基于小波倒谱模型的滤波处理技术及其在超声应力检测中的应用

在超声检测中,应力是通过超声时间延迟来测量的,而超声回波的时间延迟对噪声非常敏感。噪声严重影响应力的测量精度,噪声引起的时间延迟甚至与应力引起的时间延迟处于同一量级。而以往消噪处理都是基于盲信息,没有考虑系统响应特点,消噪效果不好,需一种有效的消噪方法。提出了一种基于倒谱模型新的滤波处理技术,从系统响应的角度,建立超声通道的小波倒谱模型,并通过对钢轨应力测试信号进行滤波处理,试验结果令人满意。

A new filtration method based on wavelet-cepstrum model and its application in ultrasonic stress detection

Ultrasonic information of stress detection is dependent on time delay.Time delay of ultrasonic wave is very sensitive to noise,so noise has great effect on detection accuracy.Time delay caused by noise is the same order as that caused by stress.But previous noise elimination method was usually based on blind information and the effect was bad because it didn't take account of the characteristics of system response,so it requires that a new effective noise elimination method be established.A new filtration method which is based on wavelet-cepstrum model was put forward,and the wavelet-cepstrum model of ultrasonic transmitting channel according to system response was set up.Its application in rail stress signal detection was satisfactory.