基于FFT及IFFT的超声波相位差检测方法

噪声和波形畸变是影响超声波相位差检测精度的两大主要原因,也是在时域中构建相位差检测模型必须要解决的关键问题。提出一种非整周数据整周期化的算法,在获取两路超声波A/D数据后,首先对非整周期采样数据进行插值,将非整周期采样的数据转化为整周期采样数据,之后将整周期数据经过快速傅里叶变换(FFT)在频域中滤除噪声和波形畸变后,再通过快速傅里叶逆变换(IFFT)由频域回到时域中进行相位差检测。实验数据表明,采用FFT及IFFT的时域-频域-时域变换的超声波相位差检测法,其相位差检测精度明显优于时域法相位差检测精度,当选用12位A/D采样超声波数据时,相位差的检测标准差<0.01°。

Method of ultrasonic phase difference detection based on FFT and IFFT

Noise and waveform distortion are two main factors which affect the accuracy of ultrasonic phase difference detection. They are also the key issues that must be solved to construct phase difference detection model in the time domain. An algorithm for transforming non-integer period sampled data into integer period sampled data is proposed. After acquiring two routes of ultrasonic A/D data, the noninteger period sampling data are interpolated, converting non-integer period sampled data into integer period sampled data. After filtering noise and waveform distortion by fast Fourier transform ( FFT), phase difference detection is carried out by Inverse fast Fourier transform (IFFT) from frequency domain to time domain. The experimental data show that the FFT and IFFT time domain-frequency domain-time domain transformations are used to detect the phase difference of ultrasound. And the accuracy of phase difference detection is obviously better than that of time domain method. When 12-bit A/D is used as sample ultrasonic data, the standard deviation of phase difference is less than 0. 01 degree.