US 12,345,520 B2
Laser and synthetic profilometry of a positive displacement motor (PDM) and ML methods, systems, and computer medium storing instructions for predicting PDM attributes
Lonnie L. Smith, II, Cleburne, TX (US); Orson M. Terrill, Houston, TX (US); and William D. Murray, Tomball, TX (US)
Assigned to Turnco LLC, Spring, TX (US)
Filed by Turnco LLC, Spring, TX (US)
Filed on Oct. 17, 2024, as Appl. No. 18/919,392.
Claims priority of provisional application 63/569,148, filed on Mar. 24, 2024.
Claims priority of provisional application 63/590,784, filed on Oct. 17, 2023.
Prior Publication US 2025/0123098 A1, Apr. 17, 2025
Int. Cl. G01B 11/24 (2006.01); G06N 3/0455 (2023.01); H02P 21/14 (2016.01)
CPC G01B 11/2441 (2013.01) [G06N 3/0455 (2023.01); H02P 21/14 (2013.01)] 30 Claims
OG exemplary drawing
 
1. A computer-implemented method for generating synthetic stator profilometric data for a positive displacement motor (PDM), comprising the steps of:
(a) providing a database of historical PDMs, wherein the historical PDMs comprise respective stators and rotors, the database comprising PDM metadata comprising stator metadata, and stator dimension data based on sets of mechanical stator minor inside diameter (SMID) measurements and radial laser profilometric (RLP) scans taken at or near a plurality of spaced axial PDM measurement locations in the respective stators of the historical PDMs;
(b) training a stator profile machine learning (SPML) model with a stator training dataset comprising the stator metadata and the stator dimension data in the database to learn stator profiles from the stator dimension data;
(c) receiving input data for a subject PDM, wherein the input data comprise input stator data, wherein the input stator data comprise input stator metadata and input mechanical SMID measurements taken at or near a plurality of spaced axial PDM measurement locations in the stator of the subject PDM in real time; and
(d) deploying the SPML model on the input mechanical SMID measurements to generate synthetic stator dimension data for the subject PDM in real time,
wherein the synthetic stator dimension data comprise:
(i) synthetic SMID values for the subject PDM selected from a vector of synthetic SMID values corresponding to the plurality of spaced axial PDM measurement locations, a mean synthetic SMID value, and a combination thereof; or
(ii) synthetic stator profilometric data, wherein the synthetic stator statistical profilometric data for the subject PDM comprise (A) one or more of local maximum, minimum, mean and/or deviation magnitude values of peaks and valleys at each one of the plurality of the spaced axial measurement locations separately, (B) one or more of overall maximum, minimum, mean and/or deviation magnitude values of peaks and valleys for the plurality of the spaced axial measurement locations collectively, and optionally (C) one or more wobble values, one or more wander values, or a combination thereof.