Research on Drill String Vibration based on the Cepstrum Analysis and Abstracting of SWD Bit Source Signals

From the vibration drilling information of the surface, we can not only provide the drill bit source signal for the SWD（Seismic While Drilling）data interpretation, estimate the condition under the well, but also get such information as the stratum character, the attrited status of the aiguille and the rotating status of the drill string. In SWD pilot data preprocessing, it is very important that effective signals are abstracted from bit. Also, noises are abstracted from rigs and machines on ground. Source signal from bit, because of the broad range of frequencies and short time duration, can be easily affected by noises from rigs and machines. In order to avoid the affection and recover the bit source signals, the source function associated with the surface record is the key approach for processing the SWD signals.

Cepstrum analysis is a nonlinear filtering technology, can change the convoluted signals in time domain to added signals in frequency domain. This method can remove the structural reverberation to abstract the source signals by selecting a window function. We discussed the cepstral filtering and abstracted the transient source signals according to the data of drill string simulated experiment.

Indoor simulation experiment verifies reliability of cepstrum analysis technology, stifles reverberation of pipe string, and obtains source signal and transfer function. On the basis of noise elimination, analyze vibration signals received by the top of drilling string using cepstrum, stifles long time cycle reverberation, highlights periodic characteristics of signals, which supplies convenience for analysis of drill bit feeble vibration and spreading characteristics. Correlate extracted drill bit source signal with ground records, which improves signal-to-noise ratio of SWD data processing. Although cepstrum can not recover exact source signals, preliminary estimates can still be given for transient source signal in accordance with amplitude and width of the recovered signals, which supplies a feasible way for SWD data processing.