Application of independent component analysis in seismic while drilling signal processing

Seismic while drilling is a technology in which the signal from the drill bit is regarded as the source, and the geological layers ahead of the drilling bit can be predicted. The signal records from the seismic while drilling have low Signal-to-Noise Ratio, and their wave fields are very complex. The vibration signal separated from the records of the acceleration sensors which is mounted at end of the drillstring is the key in the signal processing as it will be used as the source function for the autocorrelation with the surface seismic records. In this paper by using independent component analysis method 6 component from the 2 acceleration sensors were processed, by utilizing corresponding mixing matrix and waveform characteristics, the source and characteristics of every independent component was analyzed. After autocorrelating the independent component with surface seismic records, the ground roll, drill tool refraction and direct wave were estimated, and the SN Ratio was raised. The studies show that the independent component analysis method is an effective tool in solving the multi-signal overlap problem in seismic while drilling, it is a promising method to improve the data quality in seismic data processing.