Application of neural networks modeling for interpretation of acoustic logging traces

M. Lazarenko, O. Gerasimenko


The neural networks are proposed for application as a method for automatic P- and S-waves onset time-picking on sonic logging. The neural network models of acoustic emission preceding phase onset are trained and used to discriminate noise and desired signal, the last one being packets of  longitudinal  and transversal waves. The given algorithm is easily adapted to existing systems and is able to provide both processing of logging tracks in online regime and high productivity of archive materials interpretation.


neural network modelling; acoustic logging; wave arrival time; adaptive threshold level; longitudinal and transversal waves


Lazarenko M. A., Gerasimenko O. O., Ostapchuk N. N., 2006. Detection of the seismic signal using a neural network controlled. Bulletin of Kiev. Univ. Geology (is. 38/39), 47—52 (in Ukrainian).

Chen Z., Stewart R., 2005. A multi-window algorithm for real-time automatic detection and picking of P-phases of microseismic events. Conference Abstracts, CREWES, Univ. of Calgary, Canada. Ð. 14.

Chauvin Y., Rumelhart D. E., 1995. Back Propagation: Theory, Architectures and Applications. Lawrence Erlbaum Associates, 564 ð.

Guerra V., Tapia R. A., 1974. A local procedure for error detection and data smoothing. MRC Technical Summary Report 1452, Mathematics Research Center, University of Wisconsin, Madison.

Wong J., Han L., Stewart R. R., Bancro J. C., 2009. Geophysical well logs from a shallow test well and automatic time-picking on full-waveform sonic logs. CSEG Recorder 34 (4), 20—29.

Creative Commons License
Licensed under a Creative Commons Attribution 4.0 International License.

Flag Counter