Abstract: According to the vibration signal of the bit which is in the different layers with different characteristics, the delay information of the target level should be extracted. Due to the low SNR, the delay information can not be gotten by time domain waveform and the conventional cross-correlation. Wavelet transform-correlation algorithm is put forward. The reference signal and the analyzed signals will be done wavelet transformation respectively, and then cross-correlation is done in frequency domain to obtain the delay information, in order to determine the time when the bit is into the gravel medium. Both simulation and experimental research prove that compared with the traditional cross-correlation, the wavelet transform-cross correlation algorithm has a better ability to suppress the noise, and the delay information is detected accurately and clearly in real time.

Key words: wavelet transform, cross-correlation, stratum judgement, vibration signal

摘要： 钻探工程中需要依据钻头钻进不同地层产生的振动信号信息，实时判断钻头钻进目标层的时刻。时域波形图和传统的时域互相关在低信噪比时无法有效提取钻头进入不同地层的时延信息。提出了小波-互相关算法，将参考信号和待测信号分别进行小波变换后再进行频域互相关，提取待测信号相对于参考信号的时延信息，进而实时判断钻头钻进目标层的时刻。仿真和试验均表明：小波-互相关算法与传统的时域互相关相比，能实时地提取钻头钻进不同地层的时延信息，具有更好的时间定位和抑噪能力。

关键词: 小波变换, 互相关, 地层诊断, 振动信号