土体障碍物超声探测信号小波包特征分析方法

利用小波包分解方法，对土体中障碍物的超声探测信号进行分析。选用sym8小波基函数进行6层小波包分解，并截取信号能量集中的前16个频带进行细致分析，从小波包频带能量、小波包频带能量熵、小波包频带能量矩三个角度对信号进行对比。结果显示，存在障碍物位置处信号和不存在障碍物位置处信号的小波包频带能量熵大致相等，不能作为信号识别及障碍物判断的方法，而小波包频带能量和小波包频带能量矩在部分频带上差异明显，可以作为信号识别及判断的方法。

Analysis Method for Wavelet Packet Feature of Ultrasonic Detection Signals of Obstacles in Soil

The ultrasonic detection signal of obstacles in soil is analyzed by the wavelet packet decomposition method. By selecting the sym8 wavelet basis function, the 6-layer wavelet packet decomposition is conducted, and by intercepting the first 16 frequency bands of signal energy concentration, the detailed analysis is carried out. The signals are compared from three perspectives of wavelet packet frequency band energy, wavelet packet frequency band energy entropy and wavelet packet frequency band energy moment. The results show that the wavelet packet frequency band energy entropy of the signal at the location of obstacle and the signal at the location of no obstacle are roughly equal, and cannot be used as a method for signal recognition and obstacle judgment. However, the wavelet packet frequency band energy and wavelet packet frequency band energy moment have significant differences in some frequency bands, and can be used as a method for signal recognition and judgment.