US 12,072,462 B2
Machine learning enhanced borehole sonic data interpretation
Lin Liang, Belmont, MA (US); and Ting Lei, Arlington, MA (US)
Assigned to SCHLUMBERGER TECHNOLOGY CORPORATION, Sugar Land, TX (US)
Appl. No. 17/618,714
Filed by Schlumberger Technology Corporation, Sugar Land, TX (US)
PCT Filed Jun. 13, 2020, PCT No. PCT/US2020/037645
§ 371(c)(1), (2) Date Dec. 13, 2021,
PCT Pub. No. WO2020/252419, PCT Pub. Date Dec. 17, 2020.
Claims priority of provisional application 62/902,009, filed on Sep. 18, 2019.
Claims priority of provisional application 62/861,756, filed on Jun. 14, 2019.
Prior Publication US 2022/0244419 A1, Aug. 4, 2022
Int. Cl. G01V 1/50 (2006.01); G01V 1/46 (2006.01); G06N 3/08 (2023.01)
CPC G01V 1/50 (2013.01) [G01V 1/46 (2013.01); G06N 3/08 (2013.01); G01V 2210/20 (2013.01); G01V 2210/626 (2013.01)] 5 Claims
OG exemplary drawing
 
1. A method for processing borehole sonic dispersion data, comprising:
defining a frequency range;
defining a slowness range;
receiving the borehole sonic dispersion data;
categorizing the borehole sonic dispersion data into a plurality of discrete slowness-frequency points based at least in part on the defined frequency and slowness ranges;
classifying the plurality of discrete slowness-frequency points into a plurality of cluster classes;
identifying each respective cluster class in the plurality of cluster classes that exhibit large misfit residual error and removing the respective cluster class;
combining the plurality of discrete slowness-frequency points in the remaining cluster classes into a combined dataset;
inverting the combined dataset to generate a dipole dispersion curve and a refined dataset;
until a stop criterion is met, iteratively:
applying an adaptive filter to the refined dataset to identify and remove points in the refined dataset that are outliers from the dipole dispersion curve; and
inverting the refined dataset; and
filtering the refined dataset based at least in part on a predefined distance from the dipole dispersion curve.